Datasets:

yuntian-deng

/

ak-paper-selection

like 3

2402.09957

2024-02-15T14:08:08Z

On Designing Features for Condition Monitoring of Rotating Machines

Various methods for designing input features have been proposed for fault recognition in rotating machines using one-dimensional raw sensor data. The available methods are complex, rely on empirical approaches, and may differ depending on the condition monitoring data used. Therefore, this article proposes a novel algorithm to design input features that unifies the feature extraction process for different time-series sensor data. This new insight for designing/extracting input features is obtained through the lens of histogram theory. The proposed algorithm extracts discriminative input features, which are suitable for a simple classifier to deep neural network-based classifiers. The designed input features are given as input to the classifier with end-to-end training in a single framework for machine conditions recognition. The proposed scheme has been validated through three real-time datasets: a) acoustic dataset, b) CWRU vibration dataset, and c) IMS vibration dataset. The real-time results and comparative study show the effectiveness of the proposed scheme for the prediction of the machine's health states.

2402.09970

2024-02-15T14:27:58Z

Accelerating Parallel Sampling of Diffusion Models

Diffusion models have emerged as state-of-the-art generative models for image generation. However, sampling from diffusion models is usually time-consuming due to the inherent autoregressive nature of their sampling process. In this work, we propose a novel approach that accelerates the sampling of diffusion models by parallelizing the autoregressive process. Specifically, we reformulate the sampling process as solving a system of triangular nonlinear equations through fixed-point iteration. With this innovative formulation, we explore several systematic techniques to further reduce the iteration steps required by the solving process. Applying these techniques, we introduce ParaTAA, a universal and training-free parallel sampling algorithm that can leverage extra computational and memory resources to increase the sampling speed. Our experiments demonstrate that ParaTAA can decrease the inference steps required by common sequential sampling algorithms such as DDIM and DDPM by a factor of 4~14 times. Notably, when applying ParaTAA with 100 steps DDIM for Stable Diffusion, a widely-used text-to-image diffusion model, it can produce the same images as the sequential sampling in only 7 inference steps.

2402.09978

2024-02-15T14:41:55Z

Deep learning for the design of non-Hermitian topolectrical circuits

Non-Hermitian topological phases can produce some remarkable properties, compared with their Hermitian counterpart, such as the breakdown of conventional bulk-boundary correspondence and the non-Hermitian topological edge mode. Here, we introduce several algorithms with multi-layer perceptron (MLP), and convolutional neural network (CNN) in the field of deep learning, to predict the winding of eigenvalues non-Hermitian Hamiltonians. Subsequently, we use the smallest module of the periodic circuit as one unit to construct high-dimensional circuit data features. Further, we use the Dense Convolutional Network (DenseNet), a type of convolutional neural network that utilizes dense connections between layers to design a non-Hermitian topolectrical Chern circuit, as the DenseNet algorithm is more suitable for processing high-dimensional data. Our results demonstrate the effectiveness of the deep learning network in capturing the global topological characteristics of a non-Hermitian system based on training data.

2402.09984

2024-02-15T14:49:28Z

Symmetry-Breaking Augmentations for Ad Hoc Teamwork

In many collaborative settings, artificial intelligence (AI) agents must be able to adapt to new teammates that use unknown or previously unobserved strategies. While often simple for humans, this can be challenging for AI agents. For example, if an AI agent learns to drive alongside others (a training set) that only drive on one side of the road, it may struggle to adapt this experience to coordinate with drivers on the opposite side, even if their behaviours are simply flipped along the left-right symmetry. To address this we introduce symmetry-breaking augmentations (SBA), which increases diversity in the behaviour of training teammates by applying a symmetry-flipping operation. By learning a best-response to the augmented set of teammates, our agent is exposed to a wider range of behavioural conventions, improving performance when deployed with novel teammates. We demonstrate this experimentally in two settings, and show that our approach improves upon previous ad hoc teamwork results in the challenging card game Hanabi. We also propose a general metric for estimating symmetry-dependency amongst a given set of policies.

2402.10001

2024-02-15T15:06:33Z

Privacy Attacks in Decentralized Learning

Decentralized Gradient Descent (D-GD) allows a set of users to perform collaborative learning without sharing their data by iteratively averaging local model updates with their neighbors in a network graph. The absence of direct communication between non-neighbor nodes might lead to the belief that users cannot infer precise information about the data of others. In this work, we demonstrate the opposite, by proposing the first attack against D-GD that enables a user (or set of users) to reconstruct the private data of other users outside their immediate neighborhood. Our approach is based on a reconstruction attack against the gossip averaging protocol, which we then extend to handle the additional challenges raised by D-GD. We validate the effectiveness of our attack on real graphs and datasets, showing that the number of users compromised by a single or a handful of attackers is often surprisingly large. We empirically investigate some of the factors that affect the performance of the attack, namely the graph topology, the number of attackers, and their position in the graph.

2402.10046

2024-02-15T16:07:56Z

How Flawed is ECE? An Analysis via Logit Smoothing

Informally, a model is calibrated if its predictions are correct with a probability that matches the confidence of the prediction. By far the most common method in the literature for measuring calibration is the expected calibration error (ECE). Recent work, however, has pointed out drawbacks of ECE, such as the fact that it is discontinuous in the space of predictors. In this work, we ask: how fundamental are these issues, and what are their impacts on existing results? Towards this end, we completely characterize the discontinuities of ECE with respect to general probability measures on Polish spaces. We then use the nature of these discontinuities to motivate a novel continuous, easily estimated miscalibration metric, which we term Logit-Smoothed ECE (LS-ECE). By comparing the ECE and LS-ECE of pre-trained image classification models, we show in initial experiments that binned ECE closely tracks LS-ECE, indicating that the theoretical pathologies of ECE may be avoidable in practice.

2402.10082

2024-02-15T16:42:04Z

FedRDF: A Robust and Dynamic Aggregation Function against Poisoning Attacks in Federated Learning

Federated Learning (FL) represents a promising approach to typical privacy concerns associated with centralized Machine Learning (ML) deployments. Despite its well-known advantages, FL is vulnerable to security attacks such as Byzantine behaviors and poisoning attacks, which can significantly degrade model performance and hinder convergence. The effectiveness of existing approaches to mitigate complex attacks, such as median, trimmed mean, or Krum aggregation functions, has been only partially demonstrated in the case of specific attacks. Our study introduces a novel robust aggregation mechanism utilizing the Fourier Transform (FT), which is able to effectively handling sophisticated attacks without prior knowledge of the number of attackers. Employing this data technique, weights generated by FL clients are projected into the frequency domain to ascertain their density function, selecting the one exhibiting the highest frequency. Consequently, malicious clients' weights are excluded. Our proposed approach was tested against various model poisoning attacks, demonstrating superior performance over state-of-the-art aggregation methods.

2402.10142

2024-02-15T17:48:58Z

Tracking Changing Probabilities via Dynamic Learners

Consider a predictor, a learner, whose input is a stream of discrete items. The predictor's task, at every time point, is probabilistic multiclass prediction, i.e., to predict which item may occur next by outputting zero or more candidate items, each with a probability, after which the actual item is revealed and the predictor learns from this observation. To output probabilities, the predictor keeps track of the proportions of the items it has seen. The predictor has constant (limited) space and we seek efficient prediction and update techniques: The stream is unbounded, the set of items is unknown to the predictor and their totality can also grow unbounded. Moreover, there is non-stationarity: the underlying frequencies of items may change, substantially, from time to time. For instance, new items may start appearing and a few currently frequent items may cease to occur again. The predictor, being space-bounded, need only provide probabilities for those items with (currently) sufficiently high frequency, i.e., the salient items. This problem is motivated in the setting of prediction games, a self-supervised learning regime where concepts serve as both the predictors and the predictands, and the set of concepts grows over time, resulting in non-stationarities as new concepts are generated and used. We develop moving average techniques designed to respond to such non-stationarities in a timely manner, and explore their properties. One is a simple technique based on queuing of count snapshots, and another is a combination of queuing together with an extended version of sparse EMA. The latter combination supports predictand-specific dynamic learning rates. We find that this flexibility allows for a more accurate and timely convergence.

2402.10164

2024-02-15T18:09:41Z

Random features and polynomial rules

Random features models play a distinguished role in the theory of deep learning, describing the behavior of neural networks close to their infinite-width limit. In this work, we present a thorough analysis of the generalization performance of random features models for generic supervised learning problems with Gaussian data. Our approach, built with tools from the statistical mechanics of disordered systems, maps the random features model to an equivalent polynomial model, and allows us to plot average generalization curves as functions of the two main control parameters of the problem: the number of random features $N$ and the size $P$ of the training set, both assumed to scale as powers in the input dimension $D$. Our results extend the case of proportional scaling between $N$, $P$ and $D$. They are in accordance with rigorous bounds known for certain particular learning tasks and are in quantitative agreement with numerical experiments performed over many order of magnitudes of $N$ and $P$. We find good agreement also far from the asymptotic limits where $D\to \infty$ and at least one between $P/D^K$, $N/D^L$ remains finite.

2402.10177

2024-02-15T18:27:18Z

Large Scale Constrained Clustering With Reinforcement Learning

Given a network, allocating resources at clusters level, rather than at each node, enhances efficiency in resource allocation and usage. In this paper, we study the problem of finding fully connected disjoint clusters to minimize the intra-cluster distances and maximize the number of nodes assigned to the clusters, while also ensuring that no two nodes within a cluster exceed a threshold distance. While the problem can easily be formulated using a binary linear model, traditional combinatorial optimization solvers struggle when dealing with large-scale instances. We propose an approach to solve this constrained clustering problem via reinforcement learning. Our method involves training an agent to generate both feasible and (near) optimal solutions. The agent learns problem-specific heuristics, tailored to the instances encountered in this task. In the results section, we show that our algorithm finds near optimal solutions, even for large scale instances.

2402.10206

2024-02-15T18:58:18Z

Ising on the Graph: Task-specific Graph Subsampling via the Ising Model

Reducing a graph while preserving its overall structure is an important problem with many applications. Typically, the reduction approaches either remove edges (sparsification) or merge nodes (coarsening) in an unsupervised way with no specific downstream task in mind. In this paper, we present an approach for subsampling graph structures using an Ising model defined on either the nodes or edges and learning the external magnetic field of the Ising model using a graph neural network. Our approach is task-specific as it can learn how to reduce a graph for a specific downstream task in an end-to-end fashion. The utilized loss function of the task does not even have to be differentiable. We showcase the versatility of our approach on three distinct applications: image segmentation, 3D shape sparsification, and sparse approximate matrix inverse determination.

2402.10248

2024-02-15T11:09:22Z

A Data-Driven Supervised Machine Learning Approach to Estimating Global Ambient Air Pollution Concentrations With Associated Prediction Intervals

Global ambient air pollution, a transboundary challenge, is typically addressed through interventions relying on data from spatially sparse and heterogeneously placed monitoring stations. These stations often encounter temporal data gaps due to issues such as power outages. In response, we have developed a scalable, data-driven, supervised machine learning framework. This model is designed to impute missing temporal and spatial measurements, thereby generating a comprehensive dataset for pollutants including NO$_2$, O$_3$, PM$_{10}$, PM$_{2.5}$, and SO$_2$. The dataset, with a fine granularity of 0.25$^{\circ}$ at hourly intervals and accompanied by prediction intervals for each estimate, caters to a wide range of stakeholders relying on outdoor air pollution data for downstream assessments. This enables more detailed studies. Additionally, the model's performance across various geographical locations is examined, providing insights and recommendations for strategic placement of future monitoring stations to further enhance the model's accuracy.

2402.10282

2024-02-15T19:18:47Z

Information Capacity Regret Bounds for Bandits with Mediator Feedback

This work addresses the mediator feedback problem, a bandit game where the decision set consists of a number of policies, each associated with a probability distribution over a common space of outcomes. Upon choosing a policy, the learner observes an outcome sampled from its distribution and incurs the loss assigned to this outcome in the present round. We introduce the policy set capacity as an information-theoretic measure for the complexity of the policy set. Adopting the classical EXP4 algorithm, we provide new regret bounds depending on the policy set capacity in both the adversarial and the stochastic settings. For a selection of policy set families, we prove nearly-matching lower bounds, scaling similarly with the capacity. We also consider the case when the policies' distributions can vary between rounds, thus addressing the related bandits with expert advice problem, which we improve upon its prior results. Additionally, we prove a lower bound showing that exploiting the similarity between the policies is not possible in general under linear bandit feedback. Finally, for a full-information variant, we provide a regret bound scaling with the information radius of the policy set.

2402.10289

2024-02-15T19:37:39Z

Thompson Sampling in Partially Observable Contextual Bandits

Contextual bandits constitute a classical framework for decision-making under uncertainty. In this setting, the goal is to learn the arms of highest reward subject to contextual information, while the unknown reward parameters of each arm need to be learned by experimenting that specific arm. Accordingly, a fundamental problem is that of balancing exploration (i.e., pulling different arms to learn their parameters), versus exploitation (i.e., pulling the best arms to gain reward). To study this problem, the existing literature mostly considers perfectly observed contexts. However, the setting of partial context observations remains unexplored to date, despite being theoretically more general and practically more versatile. We study bandit policies for learning to select optimal arms based on the data of observations, which are noisy linear functions of the unobserved context vectors. Our theoretical analysis shows that the Thompson sampling policy successfully balances exploration and exploitation. Specifically, we establish the followings: (i) regret bounds that grow poly-logarithmically with time, (ii) square-root consistency of parameter estimation, and (iii) scaling of the regret with other quantities including dimensions and number of arms. Extensive numerical experiments with both real and synthetic data are presented as well, corroborating the efficacy of Thompson sampling. To establish the results, we introduce novel martingale techniques and concentration inequalities to address partially observed dependent random variables generated from unspecified distributions, and also leverage problem-dependent information to sharpen probabilistic bounds for time-varying suboptimality gaps. These techniques pave the road towards studying other decision-making problems with contextual information as well as partial observations.

2402.10350

2024-02-15T22:43:02Z

Large Language Models for Forecasting and Anomaly Detection: A Systematic Literature Review

This systematic literature review comprehensively examines the application of Large Language Models (LLMs) in forecasting and anomaly detection, highlighting the current state of research, inherent challenges, and prospective future directions. LLMs have demonstrated significant potential in parsing and analyzing extensive datasets to identify patterns, predict future events, and detect anomalous behavior across various domains. However, this review identifies several critical challenges that impede their broader adoption and effectiveness, including the reliance on vast historical datasets, issues with generalizability across different contexts, the phenomenon of model hallucinations, limitations within the models' knowledge boundaries, and the substantial computational resources required. Through detailed analysis, this review discusses potential solutions and strategies to overcome these obstacles, such as integrating multimodal data, advancements in learning methodologies, and emphasizing model explainability and computational efficiency. Moreover, this review outlines critical trends that are likely to shape the evolution of LLMs in these fields, including the push toward real-time processing, the importance of sustainable modeling practices, and the value of interdisciplinary collaboration. Conclusively, this review underscores the transformative impact LLMs could have on forecasting and anomaly detection while emphasizing the need for continuous innovation, ethical considerations, and practical solutions to realize their full potential.

2402.10972

2024-02-15T08:30:50Z

Modeling methodology for the accurate and prompt prediction of symptomatic events in chronic diseases

Prediction of symptomatic crises in chronic diseases allows to take decisions before the symptoms occur, such as the intake of drugs to avoid the symptoms or the activation of medical alarms. The prediction horizon is in this case an important parameter in order to fulfill the pharmacokinetics of medications, or the time response of medical services. This paper presents a study about the prediction limits of a chronic disease with symptomatic crises: the migraine. For that purpose, this work develops a methodology to build predictive migraine models and to improve these predictions beyond the limits of the initial models. The maximum prediction horizon is analyzed, and its dependency on the selected features is studied. A strategy for model selection is proposed to tackle the trade off between conservative but robust predictive models, with respect to less accurate predictions with higher horizons. The obtained results show a prediction horizon close to 40 minutes, which is in the time range of the drug pharmacokinetics. Experiments have been performed in a realistic scenario where input data have been acquired in an ambulatory clinical study by the deployment of a non-intrusive Wireless Body Sensor Network. Our results provide an effective methodology for the selection of the future horizon in the development of prediction algorithms for diseases experiencing symptomatic crises.

2402.15521

2024-02-15T18:13:41Z

HKD-SHO: A hybrid smart home system based on knowledge-based and data-driven services

A smart home is realized by setting up various services. Several methods have been proposed to create smart home services, which can be divided into knowledge-based and data-driven approaches. However, knowledge-based approaches usually require manual input from the inhabitant, which can be complicated if the physical phenomena of the concerned environment states are complex, and the inhabitant does not know how to adjust related actuators to achieve the target values of the states monitored by services. Moreover, machine learning-based data-driven approaches that we are interested in are like black boxes and cannot show the inhabitant in which situations certain services proposed certain actuators' states. To solve these problems, we propose a hybrid system called HKD-SHO (Hybrid Knowledge-based and Data-driven services based Smart HOme system), where knowledge-based and machine learning-based data-driven services are profitably integrated. The principal advantage is that it inherits the explicability of knowledge-based services and the dynamism of data-driven services. We compare HKD-SHO with several systems for creating dynamic smart home services, and the results show the better performance of HKD-SHO.

2402.17771

2024-02-15T18:49:05Z

Utilizing Machine Learning for Signal Classification and Noise Reduction in Amateur Radio

In the realm of amateur radio, the effective classification of signals and the mitigation of noise play crucial roles in ensuring reliable communication. Traditional methods for signal classification and noise reduction often rely on manual intervention and predefined thresholds, which can be labor-intensive and less adaptable to dynamic radio environments. In this paper, we explore the application of machine learning techniques for signal classification and noise reduction in amateur radio operations. We investigate the feasibility and effectiveness of employing supervised and unsupervised learning algorithms to automatically differentiate between desired signals and unwanted interference, as well as to reduce the impact of noise on received transmissions. Experimental results demonstrate the potential of machine learning approaches to enhance the efficiency and robustness of amateur radio communication systems, paving the way for more intelligent and adaptive radio solutions in the amateur radio community.

2403.05559

2024-02-15T14:12:38Z

Improving Cognitive Diagnosis Models with Adaptive Relational Graph Neural Networks

Cognitive Diagnosis (CD) algorithms receive growing research interest in intelligent education. Typically, these CD algorithms assist students by inferring their abilities (i.e., their proficiency levels on various knowledge concepts). The proficiency levels can enable further targeted skill training and personalized exercise recommendations, thereby promoting students' learning efficiency in online education. Recently, researchers have found that building and incorporating a student-exercise bipartite graph is beneficial for enhancing diagnostic performance. However, there are still limitations in their studies. On one hand, researchers overlook the heterogeneity within edges, where there can be both correct and incorrect answers. On the other hand, they disregard the uncertainty within edges, e.g., a correct answer can indicate true mastery or fortunate guessing. To address the limitations, we propose Adaptive Semantic-aware Graph-based Cognitive Diagnosis model (ASG-CD), which introduces a novel and effective way to leverage bipartite graph information in CD. Specifically, we first map students, exercises, and knowledge concepts into a latent representation space and combine these latent representations to obtain student abilities and exercise difficulties. After that, we propose a Semantic-aware Graph Neural Network Layer to address edge heterogeneity. This layer splits the original bipartite graph into two subgraphs according to edge semantics, and aggregates information based on these two subgraphs separately. To mitigate the impact of edge uncertainties, we propose an Adaptive Edge Differentiation Layer that dynamically differentiates edges, followed by keeping reliable edges and filtering out uncertain edges. Extensive experiments on three real-world datasets have demonstrated the effectiveness of ASG-CD.

2403.15394

2024-02-15T14:56:00Z

"Model Cards for Model Reporting" in 2024: Reclassifying Category of Ethical Considerations in Terms of Trustworthiness and Risk Management

In 2019, the paper entitled "Model Cards for Model Reporting" introduced a new tool for documenting model performance and encouraged the practice of transparent reporting for a defined list of categories. One of the categories detailed in that paper is ethical considerations, which includes the subcategories of data, human life, mitigations, risks and harms, and use cases. We propose to reclassify this category in the original model card due to the recent maturing of the field known as trustworthy AI, a term which analyzes whether the algorithmic properties of the model indicate that the AI system is deserving of trust from its stakeholders. In our examination of trustworthy AI, we highlight three respected organizations - the European Commission's High-Level Expert Group on AI, the OECD, and the U.S.-based NIST - that have written guidelines on various aspects of trustworthy AI. These recent publications converge on numerous characteristics of the term, including accountability, explainability, fairness, privacy, reliability, robustness, safety, security, and transparency, while recognizing that the implementation of trustworthy AI varies by context. Our reclassification of the original model-card category known as ethical considerations involves a two-step process: 1) adding a new category known as trustworthiness, where the subcategories will be derived from the discussion of trustworthy AI in our paper, and 2) maintaining the subcategories of ethical considerations under a renamed category known as risk environment and risk management, a title which we believe better captures today's understanding of the essence of these topics. We hope that this reclassification will further the goals of the original paper and continue to prompt those releasing trained models to accompany these models with documentation that will assist in the evaluation of their algorithmic properties.

2402.09657

2024-02-15T01:50:46Z

Digital versus Analog Transmissions for Federated Learning over Wireless Networks

In this paper, we quantitatively compare these two effective communication schemes, i.e., digital and analog ones, for wireless federated learning (FL) over resource-constrained networks, highlighting their essential differences as well as their respective application scenarios. We first examine both digital and analog transmission methods, together with a unified and fair comparison scheme under practical constraints. A universal convergence analysis under various imperfections is established for FL performance evaluation in wireless networks. These analytical results reveal that the fundamental difference between the two paradigms lies in whether communication and computation are jointly designed or not. The digital schemes decouple the communication design from specific FL tasks, making it difficult to support simultaneous uplink transmission of massive devices with limited bandwidth. In contrast, the analog communication allows over-the-air computation (AirComp), thus achieving efficient spectrum utilization. However, computation-oriented analog transmission reduces power efficiency, and its performance is sensitive to computational errors. Finally, numerical simulations are conducted to verify these theoretical observations.

2402.09695

2024-02-15T04:08:49Z

Reward Poisoning Attack Against Offline Reinforcement Learning

We study the problem of reward poisoning attacks against general offline reinforcement learning with deep neural networks for function approximation. We consider a black-box threat model where the attacker is completely oblivious to the learning algorithm and its budget is limited by constraining both the amount of corruption at each data point, and the total perturbation. We propose an attack strategy called `policy contrast attack'. The high-level idea is to make some low-performing policies appear as high-performing while making high-performing policies appear as low-performing. To the best of our knowledge, we propose the first black-box reward poisoning attack in the general offline RL setting. We provide theoretical insights on the attack design and empirically show that our attack is efficient against current state-of-the-art offline RL algorithms in different kinds of learning datasets.

2402.09710

2024-02-15T05:06:53Z

Preserving Data Privacy for ML-driven Applications in Open Radio Access Networks

Deep learning offers a promising solution to improve spectrum access techniques by utilizing data-driven approaches to manage and share limited spectrum resources for emerging applications. For several of these applications, the sensitive wireless data (such as spectrograms) are stored in a shared database or multistakeholder cloud environment and are therefore prone to privacy leaks. This paper aims to address such privacy concerns by examining the representative case study of shared database scenarios in 5G Open Radio Access Network (O-RAN) networks where we have a shared database within the near-real-time (near-RT) RAN intelligent controller. We focus on securing the data that can be used by machine learning (ML) models for spectrum sharing and interference mitigation applications without compromising the model and network performances. The underlying idea is to leverage a (i) Shuffling-based learnable encryption technique to encrypt the data, following which, (ii) employ a custom Vision transformer (ViT) as the trained ML model that is capable of performing accurate inferences on such encrypted data. The paper offers a thorough analysis and comparisons with analogous convolutional neural networks (CNN) as well as deeper architectures (such as ResNet-50) as baselines. Our experiments showcase that the proposed approach significantly outperforms the baseline CNN with an improvement of 24.5% and 23.9% for the percent accuracy and F1-Score respectively when operated on encrypted data. Though deeper ResNet-50 architecture is obtained as a slightly more accurate model, with an increase of 4.4%, the proposed approach boasts a reduction of parameters by 99.32%, and thus, offers a much-improved prediction time by nearly 60%.

2402.09715

2024-02-15T05:19:53Z

DPBalance: Efficient and Fair Privacy Budget Scheduling for Federated Learning as a Service

Federated learning (FL) has emerged as a prevalent distributed machine learning scheme that enables collaborative model training without aggregating raw data. Cloud service providers further embrace Federated Learning as a Service (FLaaS), allowing data analysts to execute their FL training pipelines over differentially-protected data. Due to the intrinsic properties of differential privacy, the enforced privacy level on data blocks can be viewed as a privacy budget that requires careful scheduling to cater to diverse training pipelines. Existing privacy budget scheduling studies prioritize either efficiency or fairness individually. In this paper, we propose DPBalance, a novel privacy budget scheduling mechanism that jointly optimizes both efficiency and fairness. We first develop a comprehensive utility function incorporating data analyst-level dominant shares and FL-specific performance metrics. A sequential allocation mechanism is then designed using the Lagrange multiplier method and effective greedy heuristics. We theoretically prove that DPBalance satisfies Pareto Efficiency, Sharing Incentive, Envy-Freeness, and Weak Strategy Proofness. We also theoretically prove the existence of a fairness-efficiency tradeoff in privacy budgeting. Extensive experiments demonstrate that DPBalance outperforms state-of-the-art solutions, achieving an average efficiency improvement of $1.44\times \sim 3.49 \times$, and an average fairness improvement of $1.37\times \sim 24.32 \times$.

2402.09735

2024-02-15T06:22:50Z

DFORM: Diffeomorphic vector field alignment for assessing dynamics across learned models

Dynamical system models such as Recurrent Neural Networks (RNNs) have become increasingly popular as hypothesis-generating tools in scientific research. Evaluating the dynamics in such networks is key to understanding their learned generative mechanisms. However, comparison of learned dynamics across models is challenging due to their inherent nonlinearity and because a priori there is no enforced equivalence of their coordinate systems. Here, we propose the DFORM (Diffeomorphic vector field alignment for comparing dynamics across learned models) framework. DFORM learns a nonlinear coordinate transformation which provides a continuous, maximally one-to-one mapping between the trajectories of learned models, thus approximating a diffeomorphism between them. The mismatch between DFORM-transformed vector fields defines the orbital similarity between two models, thus providing a generalization of the concepts of smooth orbital and topological equivalence. As an example, we apply DFORM to models trained on a canonical neuroscience task, showing that learned dynamics may be functionally similar, despite overt differences in attractor landscapes.

2402.09754

2024-02-15T07:08:11Z

Robust SVD Made Easy: A fast and reliable algorithm for large-scale data analysis

The singular value decomposition (SVD) is a crucial tool in machine learning and statistical data analysis. However, it is highly susceptible to outliers in the data matrix. Existing robust SVD algorithms often sacrifice speed for robustness or fail in the presence of only a few outliers. This study introduces an efficient algorithm, called Spherically Normalized SVD, for robust SVD approximation that is highly insensitive to outliers, computationally scalable, and provides accurate approximations of singular vectors. The proposed algorithm achieves remarkable speed by utilizing only two applications of a standard reduced-rank SVD algorithm to appropriately scaled data, significantly outperforming competing algorithms in computation times. To assess the robustness of the approximated singular vectors and their subspaces against data contamination, we introduce new notions of breakdown points for matrix-valued input, including row-wise, column-wise, and block-wise breakdown points. Theoretical and empirical analyses demonstrate that our algorithm exhibits higher breakdown points compared to standard SVD and its modifications. We empirically validate the effectiveness of our approach in applications such as robust low-rank approximation and robust principal component analysis of high-dimensional microarray datasets. Overall, our study presents a highly efficient and robust solution for SVD approximation that overcomes the limitations of existing algorithms in the presence of outliers.

2402.09761

2024-02-15T07:23:34Z

A Framework For Gait-Based User Demography Estimation Using Inertial Sensors

Human gait has been shown to provide crucial motion cues for various applications. Recognizing patterns in human gait has been widely adopted in various application areas such as security, virtual reality gaming, medical rehabilitation, and ailment identification. Furthermore, wearable inertial sensors have been widely used for not only recording gait but also to predict users' demography. Machine Learning techniques such as deep learning, combined with inertial sensor signals, have shown promising results in recognizing patterns in human gait and estimate users' demography. However, the black-box nature of such deep learning models hinders the researchers from uncovering the reasons behind the model's predictions. Therefore, we propose leveraging deep learning and Layer-Wise Relevance Propagation (LRP) to identify the important variables that play a vital role in identifying the users' demography such as age and gender. To assess the efficacy of this approach we train a deep neural network model on a large sensor-based gait dataset consisting of 745 subjects to identify users' age and gender. Using LRP we identify the variables relevant for characterizing the gait patterns. Thus, we enable interpretation of non-linear ML models which are experts in identifying the users' demography based on inertial signals. We believe this approach can not only provide clinicians information about the gait parameters relevant to age and gender but also can be expanded to analyze and diagnose gait disorders.

2402.09766

2024-02-15T07:35:52Z

From Variability to Stability: Advancing RecSys Benchmarking Practices

In the rapidly evolving domain of Recommender Systems (RecSys), new algorithms frequently claim state-of-the-art performance based on evaluations over a limited set of arbitrarily selected datasets. However, this approach may fail to holistically reflect their effectiveness due to the significant impact of dataset characteristics on algorithm performance. Addressing this deficiency, this paper introduces a novel benchmarking methodology to facilitate a fair and robust comparison of RecSys algorithms, thereby advancing evaluation practices. By utilizing a diverse set of $30$ open datasets, including two introduced in this work, and evaluating $11$ collaborative filtering algorithms across $9$ metrics, we critically examine the influence of dataset characteristics on algorithm performance. We further investigate the feasibility of aggregating outcomes from multiple datasets into a unified ranking. Through rigorous experimental analysis, we validate the reliability of our methodology under the variability of datasets, offering a benchmarking strategy that balances quality and computational demands. This methodology enables a fair yet effective means of evaluating RecSys algorithms, providing valuable guidance for future research endeavors.

2402.09796

2024-02-15T08:51:49Z

Closed-form Filtering for Non-linear Systems

Sequential Bayesian Filtering aims to estimate the current state distribution of a Hidden Markov Model, given the past observations. The problem is well-known to be intractable for most application domains, except in notable cases such as the tabular setting or for linear dynamical systems with gaussian noise. In this work, we propose a new class of filters based on Gaussian PSD Models, which offer several advantages in terms of density approximation and computational efficiency. We show that filtering can be efficiently performed in closed form when transitions and observations are Gaussian PSD Models. When the transition and observations are approximated by Gaussian PSD Models, we show that our proposed estimator enjoys strong theoretical guarantees, with estimation error that depends on the quality of the approximation and is adaptive to the regularity of the transition probabilities. In particular, we identify regimes in which our proposed filter attains a TV $\epsilon$-error with memory and computational complexity of $O(\epsilon^{-1})$ and $O(\epsilon^{-3/2})$ respectively, including the offline learning step, in contrast to the $O(\epsilon^{-2})$ complexity of sampling methods such as particle filtering.

2402.09807

2024-02-15T09:13:59Z

Two trust region type algorithms for solving nonconvex-strongly concave minimax problems

In this paper, we propose a Minimax Trust Region (MINIMAX-TR) algorithm and a Minimax Trust Region Algorithm with Contractions and Expansions(MINIMAX-TRACE) algorithm for solving nonconvex-strongly concave minimax problems. Both algorithms can find an $(\epsilon, \sqrt{\epsilon})$-second order stationary point(SSP) within $\mathcal{O}(\epsilon^{-1.5})$ iterations, which matches the best well known iteration complexity.

2402.09821

2024-02-15T09:36:36Z

Diffusion Models for Audio Restoration

With the development of audio playback devices and fast data transmission, the demand for high sound quality is rising, for both entertainment and communications. In this quest for better sound quality, challenges emerge from distortions and interferences originating at the recording side or caused by an imperfect transmission pipeline. To address this problem, audio restoration methods aim to recover clean sound signals from the corrupted input data. We present here audio restoration algorithms based on diffusion models, with a focus on speech enhancement and music restoration tasks. Traditional approaches, often grounded in handcrafted rules and statistical heuristics, have shaped our understanding of audio signals. In the past decades, there has been a notable shift towards data-driven methods that exploit the modeling capabilities of deep neural networks (DNNs). Deep generative models, and among them diffusion models, have emerged as powerful techniques for learning complex data distributions. However, relying solely on DNN-based learning approaches carries the risk of reducing interpretability, particularly when employing end-to-end models. Nonetheless, data-driven approaches allow more flexibility in comparison to statistical model-based frameworks whose performance depends on distributional and statistical assumptions that can be difficult to guarantee. Here, we aim to show that diffusion models can combine the best of both worlds and offer the opportunity to design audio restoration algorithms with a good degree of interpretability and a remarkable performance in terms of sound quality.

2402.09830

2024-02-15T09:48:20Z

Utilizing GANs for Fraud Detection: Model Training with Synthetic Transaction Data

Anomaly detection is a critical challenge across various research domains, aiming to identify instances that deviate from normal data distributions. This paper explores the application of Generative Adversarial Networks (GANs) in fraud detection, comparing their advantages with traditional methods. GANs, a type of Artificial Neural Network (ANN), have shown promise in modeling complex data distributions, making them effective tools for anomaly detection. The paper systematically describes the principles of GANs and their derivative models, emphasizing their application in fraud detection across different datasets. And by building a collection of adversarial verification graphs, we will effectively prevent fraud caused by bots or automated systems and ensure that the users in the transaction are real. The objective of the experiment is to design and implement a fake face verification code and fraud detection system based on Generative Adversarial network (GANs) algorithm to enhance the security of the transaction process.The study demonstrates the potential of GANs in enhancing transaction security through deep learning techniques.

2402.09846

2024-02-15T10:05:18Z

A Deep Learning Approach to Radar-based QPE

In this study, we propose a volume-to-point framework for quantitative precipitation estimation (QPE) based on the Quantitative Precipitation Estimation and Segregation Using Multiple Sensor (QPESUMS) Mosaic Radar data set. With a data volume consisting of the time series of gridded radar reflectivities over the Taiwan area, we used machine learning algorithms to establish a statistical model for QPE in weather stations. The model extracts spatial and temporal features from the input data volume and then associates these features with the location-specific precipitations. In contrast to QPE methods based on the Z-R relation, we leverage the machine learning algorithms to automatically detect the evolution and movement of weather systems and associate these patterns to a location with specific topographic attributes. Specifically, we evaluated this framework with the hourly precipitation data of 45 weather stations in Taipei during 2013-2016. In comparison to the operational QPE scheme used by the Central Weather Bureau, the volume-to-point framework performed comparably well in general cases and excelled in detecting heavy-rainfall events. By using the current results as the reference benchmark, the proposed method can integrate the heterogeneous data sources and potentially improve the forecast in extreme precipitation scenarios.

2402.09941

2024-02-15T13:41:23Z

FedLion: Faster Adaptive Federated Optimization with Fewer Communication

In Federated Learning (FL), a framework to train machine learning models across distributed data, well-known algorithms like FedAvg tend to have slow convergence rates, resulting in high communication costs during training. To address this challenge, we introduce FedLion, an adaptive federated optimization algorithm that seamlessly incorporates key elements from the recently proposed centralized adaptive algorithm, Lion (Chen et al. 2o23), into the FL framework. Through comprehensive evaluations on two widely adopted FL benchmarks, we demonstrate that FedLion outperforms previous state-of-the-art adaptive algorithms, including FAFED (Wu et al. 2023) and FedDA. Moreover, thanks to the use of signed gradients in local training, FedLion substantially reduces data transmission requirements during uplink communication when compared to existing adaptive algorithms, further reducing communication costs. Last but not least, this work also includes a novel theoretical analysis, showcasing that FedLion attains faster convergence rate than established FL algorithms like FedAvg.

2402.09992

2024-02-15T14:55:38Z

Risk-Sensitive Soft Actor-Critic for Robust Deep Reinforcement Learning under Distribution Shifts

We study the robustness of deep reinforcement learning algorithms against distribution shifts within contextual multi-stage stochastic combinatorial optimization problems from the operations research domain. In this context, risk-sensitive algorithms promise to learn robust policies. While this field is of general interest to the reinforcement learning community, most studies up-to-date focus on theoretical results rather than real-world performance. With this work, we aim to bridge this gap by formally deriving a novel risk-sensitive deep reinforcement learning algorithm while providing numerical evidence for its efficacy. Specifically, we introduce discrete Soft Actor-Critic for the entropic risk measure by deriving a version of the Bellman equation for the respective Q-values. We establish a corresponding policy improvement result and infer a practical algorithm. We introduce an environment that represents typical contextual multi-stage stochastic combinatorial optimization problems and perform numerical experiments to empirically validate our algorithm's robustness against realistic distribution shifts, without compromising performance on the training distribution. We show that our algorithm is superior to risk-neutral Soft Actor-Critic as well as to two benchmark approaches for robust deep reinforcement learning. Thereby, we provide the first structured analysis on the robustness of reinforcement learning under distribution shifts in the realm of contextual multi-stage stochastic combinatorial optimization problems.

2402.10028

2024-02-15T15:48:55Z

Diffusion Models Meet Contextual Bandits with Large Action Spaces

Efficient exploration is a key challenge in contextual bandits due to the large size of their action space, where uninformed exploration can result in computational and statistical inefficiencies. Fortunately, the rewards of actions are often correlated and this can be leveraged to explore them efficiently. In this work, we capture such correlations using pre-trained diffusion models; upon which we design diffusion Thompson sampling (dTS). Both theoretical and algorithmic foundations are developed for dTS, and empirical evaluation also shows its favorable performance.

2402.10036

2024-02-15T15:59:59Z

Predictive Linear Online Tracking for Unknown Targets

In this paper, we study the problem of online tracking in linear control systems, where the objective is to follow a moving target. Unlike classical tracking control, the target is unknown, non-stationary, and its state is revealed sequentially, thus, fitting the framework of online non-stochastic control. We consider the case of quadratic costs and propose a new algorithm, called predictive linear online tracking (PLOT). The algorithm uses recursive least squares with exponential forgetting to learn a time-varying dynamic model of the target. The learned model is used in the optimal policy under the framework of receding horizon control. We show the dynamic regret of PLOT scales with $\mathcal{O}(\sqrt{TV_T})$, where $V_T$ is the total variation of the target dynamics and $T$ is the time horizon. Unlike prior work, our theoretical results hold for non-stationary targets. We implement PLOT on a real quadrotor and provide open-source software, thus, showcasing one of the first successful applications of online control methods on real hardware.

2402.10115

2024-02-15T17:10:27Z

Generating Visual Stimuli from EEG Recordings using Transformer-encoder based EEG encoder and GAN

In this study, we tackle a modern research challenge within the field of perceptual brain decoding, which revolves around synthesizing images from EEG signals using an adversarial deep learning framework. The specific objective is to recreate images belonging to various object categories by leveraging EEG recordings obtained while subjects view those images. To achieve this, we employ a Transformer-encoder based EEG encoder to produce EEG encodings, which serve as inputs to the generator component of the GAN network. Alongside the adversarial loss, we also incorporate perceptual loss to enhance the quality of the generated images.

2402.10135

2024-02-15T17:38:32Z

Benchmarking federated strategies in Peer-to-Peer Federated learning for biomedical data

The increasing requirements for data protection and privacy has attracted a huge research interest on distributed artificial intelligence and specifically on federated learning, an emerging machine learning approach that allows the construction of a model between several participants who hold their own private data. In the initial proposal of federated learning the architecture was centralised and the aggregation was done with federated averaging, meaning that a central server will orchestrate the federation using the most straightforward averaging strategy. This research is focused on testing different federated strategies in a peer-to-peer environment. The authors propose various aggregation strategies for federated learning, including weighted averaging aggregation, using different factors and strategies based on participant contribution. The strategies are tested with varying data sizes to identify the most robust ones. This research tests the strategies with several biomedical datasets and the results of the experiments show that the accuracy-based weighted average outperforms the classical federated averaging method.

2402.10145

2024-02-15T17:49:50Z

A chaotic maps-based privacy-preserving distributed deep learning for incomplete and Non-IID datasets

Federated Learning is a machine learning approach that enables the training of a deep learning model among several participants with sensitive data that wish to share their own knowledge without compromising the privacy of their data. In this research, the authors employ a secured Federated Learning method with an additional layer of privacy and proposes a method for addressing the non-IID challenge. Moreover, differential privacy is compared with chaotic-based encryption as layer of privacy. The experimental approach assesses the performance of the federated deep learning model with differential privacy using both IID and non-IID data. In each experiment, the Federated Learning process improves the average performance metrics of the deep neural network, even in the case of non-IID data.

2402.10186

2024-02-15T18:41:35Z

Self-consistent Validation for Machine Learning Electronic Structure

Machine learning has emerged as a significant approach to efficiently tackle electronic structure problems. Despite its potential, there is less guarantee for the model to generalize to unseen data that hinders its application in real-world scenarios. To address this issue, a technique has been proposed to estimate the accuracy of the predictions. This method integrates machine learning with self-consistent field methods to achieve both low validation cost and interpret-ability. This, in turn, enables exploration of the model's ability with active learning and instills confidence in its integration into real-world studies.

2402.10211

2024-02-15T18:59:43Z

Hierarchical State Space Models for Continuous Sequence-to-Sequence Modeling

Reasoning from sequences of raw sensory data is a ubiquitous problem across fields ranging from medical devices to robotics. These problems often involve using long sequences of raw sensor data (e.g. magnetometers, piezoresistors) to predict sequences of desirable physical quantities (e.g. force, inertial measurements). While classical approaches are powerful for locally-linear prediction problems, they often fall short when using real-world sensors. These sensors are typically non-linear, are affected by extraneous variables (e.g. vibration), and exhibit data-dependent drift. For many problems, the prediction task is exacerbated by small labeled datasets since obtaining ground-truth labels requires expensive equipment. In this work, we present Hierarchical State-Space Models (HiSS), a conceptually simple, new technique for continuous sequential prediction. HiSS stacks structured state-space models on top of each other to create a temporal hierarchy. Across six real-world sensor datasets, from tactile-based state prediction to accelerometer-based inertial measurement, HiSS outperforms state-of-the-art sequence models such as causal Transformers, LSTMs, S4, and Mamba by at least 23% on MSE. Our experiments further indicate that HiSS demonstrates efficient scaling to smaller datasets and is compatible with existing data-filtering techniques. Code, datasets and videos can be found on https://hiss-csp.github.io.

2402.10310

2024-02-15T20:21:40Z

Interpretable Generative Adversarial Imitation Learning

Imitation learning methods have demonstrated considerable success in teaching autonomous systems complex tasks through expert demonstrations. However, a limitation of these methods is their lack of interpretability, particularly in understanding the specific task the learning agent aims to accomplish. In this paper, we propose a novel imitation learning method that combines Signal Temporal Logic (STL) inference and control synthesis, enabling the explicit representation of the task as an STL formula. This approach not only provides a clear understanding of the task but also allows for the incorporation of human knowledge and adaptation to new scenarios through manual adjustments of the STL formulae. Additionally, we employ a Generative Adversarial Network (GAN)-inspired training approach for both the inference and the control policy, effectively narrowing the gap between the expert and learned policies. The effectiveness of our algorithm is demonstrated through two case studies, showcasing its practical applicability and adaptability.

2402.10974

2024-02-15T14:39:58Z

On the Cross-Dataset Generalization of Machine Learning for Network Intrusion Detection

Network Intrusion Detection Systems (NIDS) are a fundamental tool in cybersecurity. Their ability to generalize across diverse networks is a critical factor in their effectiveness and a prerequisite for real-world applications. In this study, we conduct a comprehensive analysis on the generalization of machine-learning-based NIDS through an extensive experimentation in a cross-dataset framework. We employ four machine learning classifiers and utilize four datasets acquired from different networks: CIC-IDS-2017, CSE-CIC-IDS2018, LycoS-IDS2017, and LycoS-Unicas-IDS2018. Notably, the last dataset is a novel contribution, where we apply corrections based on LycoS-IDS2017 to the well-known CSE-CIC-IDS2018 dataset. The results show nearly perfect classification performance when the models are trained and tested on the same dataset. However, when training and testing the models in a cross-dataset fashion, the classification accuracy is largely commensurate with random chance except for a few combinations of attacks and datasets. We employ data visualization techniques in order to provide valuable insights on the patterns in the data. Our analysis unveils the presence of anomalies in the data that directly hinder the classifiers capability to generalize the learned knowledge to new scenarios. This study enhances our comprehension of the generalization capabilities of machine-learning-based NIDS, highlighting the significance of acknowledging data heterogeneity.

2402.10981

2024-02-15T22:51:27Z

Stuck-at Faults in ReRAM Neuromorphic Circuit Array and their Correction through Machine Learning

In this paper, we study the inference accuracy of the Resistive Random Access Memory (ReRAM) neuromorphic circuit due to stuck-at faults (stuck-on, stuck-off, and stuck at a certain resistive value). A simulation framework using Python is used to perform supervised machine learning (neural network with 3 hidden layers, 1 input layer, and 1 output layer) of handwritten digits and construct a corresponding fully analog neuromorphic circuit (4 synaptic arrays) simulated by Spectre. A generic 45nm Process Development Kit (PDK) was used. We study the difference in the inference accuracy degradation due to stuck-on and stuck-off defects. Various defect patterns are studied including circular, ring, row, column, and circular-complement defects. It is found that stuck-on and stuck-off defects have a similar effect on inference accuracy. However, it is also found that if there is a spatial defect variation across the columns, the inference accuracy may be degraded significantly. We also propose a machine learning (ML) strategy to recover the inference accuracy degradation due to stuck-at faults. The inference accuracy is improved from 48% to 85% in a defective neuromorphic circuit.

2402.10982

2024-02-15T23:08:18Z

mshw, a forecasting library to predict short-term electricity demand based on multiple seasonal Holt-Winters

Transmission system operators have a growing need for more accurate forecasting of electricity demand. Current electricity systems largely require demand forecasting so that the electricity market establishes electricity prices as well as the programming of production units. The companies that are part of the electrical system use exclusive software to obtain predictions, based on the use of time series and prediction tools, whether statistical or artificial intelligence. However, the most common form of prediction is based on hybrid models that use both technologies. In any case, it is software with a complicated structure, with a large number of associated variables and that requires a high computational load to make predictions. The predictions they can offer are not much better than those that simple models can offer. In this paper we present a MATLAB toolbox created for the prediction of electrical demand. The toolbox implements multiple seasonal Holt-Winters exponential smoothing models and neural network models. The models used include the use of discrete interval mobile seasonalities (DIMS) to improve forecasting on special days. Additionally, the results of its application in various electrical systems in Europe are shown, where the results obtained can be seen. The use of this library opens a new avenue of research for the use of models with discrete and complex seasonalities in other fields of application.

2403.03222

2024-02-15T01:52:44Z

Knowledge-guided EEG Representation Learning

Self-supervised learning has produced impressive results in multimedia domains of audio, vision and speech. This paradigm is equally, if not more, relevant for the domain of biosignals, owing to the scarcity of labelled data in such scenarios. The ability to leverage large-scale unlabelled data to learn robust representations could help improve the performance of numerous inference tasks on biosignals. Given the inherent domain differences between multimedia modalities and biosignals, the established objectives for self-supervised learning may not translate well to this domain. Hence, there is an unmet need to adapt these methods to biosignal analysis. In this work we propose a self-supervised model for EEG, which provides robust performance and remarkable parameter efficiency by using state space-based deep learning architecture. We also propose a novel knowledge-guided pre-training objective that accounts for the idiosyncrasies of the EEG signal. The results indicate improved embedding representation learning and downstream performance compared to prior works on exemplary tasks. Also, the proposed objective significantly reduces the amount of pre-training data required to obtain performance equivalent to prior works.

2403.18923

2024-02-15T20:27:33Z

Nature-Guided Cognitive Evolution for Predicting Dissolved Oxygen Concentrations in North Temperate Lakes

Predicting dissolved oxygen (DO) concentrations in north temperate lakes requires a comprehensive study of phenological patterns across various ecosystems, which highlights the significance of selecting phenological features and feature interactions. Process-based models are limited by partial process knowledge or oversimplified feature representations, while machine learning models face challenges in efficiently selecting relevant feature interactions for different lake types and tasks, especially under the infrequent nature of DO data collection. In this paper, we propose a Nature-Guided Cognitive Evolution (NGCE) strategy, which represents a multi-level fusion of adaptive learning with natural processes. Specifically, we utilize metabolic process-based models to generate simulated DO labels. Using these simulated labels, we implement a multi-population cognitive evolutionary search, where models, mirroring natural organisms, adaptively evolve to select relevant feature interactions within populations for different lake types and tasks. These models are not only capable of undergoing crossover and mutation mechanisms within intra-populations but also, albeit infrequently, engage in inter-population crossover. The second stage involves refining these models by retraining them with real observed labels. We have tested the performance of our NGCE strategy in predicting daily DO concentrations across a wide range of lakes in the Midwest, USA. These lakes, varying in size, depth, and trophic status, represent a broad spectrum of north temperate lakes. Our findings demonstrate that NGCE not only produces accurate predictions with few observed labels but also, through gene maps of models, reveals sophisticated phenological patterns of different lakes.

2402.10065

2024-02-15T16:30:55Z

How Much Does Each Datapoint Leak Your Privacy? Quantifying the Per-datum Membership Leakage

We study the per-datum Membership Inference Attacks (MIAs), where an attacker aims to infer whether a fixed target datum has been included in the input dataset of an algorithm and thus, violates privacy. First, we define the membership leakage of a datum as the advantage of the optimal adversary targeting to identify it. Then, we quantify the per-datum membership leakage for the empirical mean, and show that it depends on the Mahalanobis distance between the target datum and the data-generating distribution. We further assess the effect of two privacy defences, i.e. adding Gaussian noise and sub-sampling. We quantify exactly how both of them decrease the per-datum membership leakage. Our analysis builds on a novel proof technique that combines an Edgeworth expansion of the likelihood ratio test and a Lindeberg-Feller central limit theorem. Our analysis connects the existing likelihood ratio and scalar product attacks, and also justifies different canary selection strategies used in the privacy auditing literature. Finally, our experiments demonstrate the impacts of the leakage score, the sub-sampling ratio and the noise scale on the per-datum membership leakage as indicated by the theory.

2402.10127

2024-02-15T17:31:19Z

Nonlinear spiked covariance matrices and signal propagation in deep neural networks

Many recent works have studied the eigenvalue spectrum of the Conjugate Kernel (CK) defined by the nonlinear feature map of a feedforward neural network. However, existing results only establish weak convergence of the empirical eigenvalue distribution, and fall short of providing precise quantitative characterizations of the ''spike'' eigenvalues and eigenvectors that often capture the low-dimensional signal structure of the learning problem. In this work, we characterize these signal eigenvalues and eigenvectors for a nonlinear version of the spiked covariance model, including the CK as a special case. Using this general result, we give a quantitative description of how spiked eigenstructure in the input data propagates through the hidden layers of a neural network with random weights. As a second application, we study a simple regime of representation learning where the weight matrix develops a rank-one signal component over training and characterize the alignment of the target function with the spike eigenvector of the CK on test data.

2402.10168

2024-02-15T18:11:02Z

DeepSRGM -- Sequence Classification and Ranking in Indian Classical Music with Deep Learning

A vital aspect of Indian Classical Music (ICM) is Raga, which serves as a melodic framework for compositions and improvisations alike. Raga Recognition is an important music information retrieval task in ICM as it can aid numerous downstream applications ranging from music recommendations to organizing huge music collections. In this work, we propose a deep learning based approach to Raga recognition. Our approach employs efficient pre possessing and learns temporal sequences in music data using Long Short Term Memory based Recurrent Neural Networks (LSTM-RNN). We train and test the network on smaller sequences sampled from the original audio while the final inference is performed on the audio as a whole. Our method achieves an accuracy of 88.1% and 97 % during inference on the Comp Music Carnatic dataset and its 10 Raga subset respectively making it the state-of-the-art for the Raga recognition task. Our approach also enables sequence ranking which aids us in retrieving melodic patterns from a given music data base that are closely related to the presented query sequence.

2402.10252

2024-02-15T16:16:30Z

Online Control of Linear Systems with Unbounded and Degenerate Noise

This paper investigates the problem of controlling a linear system under possibly unbounded and degenerate noise with unknown cost functions, known as an online control problem. In contrast to the existing work, which assumes the boundedness of noise, we reveal that for convex costs, an $ \widetilde{O}(\sqrt{T}) $ regret bound can be achieved even for unbounded noise, where $ T $ denotes the time horizon. Moreover, when the costs are strongly convex, we establish an $ O({\rm poly} (\log T)) $ regret bound without the assumption that noise covariance is non-degenerate, which has been required in the literature. The key ingredient in removing the rank assumption on noise is a system transformation associated with the noise covariance. This simultaneously enables the parameter reduction of an online control algorithm.

2402.10283

2024-02-15T19:19:54Z

Backdoor Attack against One-Class Sequential Anomaly Detection Models

Deep anomaly detection on sequential data has garnered significant attention due to the wide application scenarios. However, deep learning-based models face a critical security threat - their vulnerability to backdoor attacks. In this paper, we explore compromising deep sequential anomaly detection models by proposing a novel backdoor attack strategy. The attack approach comprises two primary steps, trigger generation and backdoor injection. Trigger generation is to derive imperceptible triggers by crafting perturbed samples from the benign normal data, of which the perturbed samples are still normal. The backdoor injection is to properly inject the backdoor triggers to comprise the model only for the samples with triggers. The experimental results demonstrate the effectiveness of our proposed attack strategy by injecting backdoors on two well-established one-class anomaly detection models.

2402.10360

2024-02-15T23:10:45Z

Learnability is a Compact Property

Recent work on learning has yielded a striking result: the learnability of various problems can be undecidable, or independent of the standard ZFC axioms of set theory. Furthermore, the learnability of such problems can fail to be a property of finite character: informally, it cannot be detected by examining finite projections of the problem. On the other hand, learning theory abounds with notions of dimension that characterize learning and consider only finite restrictions of the problem, i.e., are properties of finite character. How can these results be reconciled? More precisely, which classes of learning problems are vulnerable to logical undecidability, and which are within the grasp of finite characterizations? We demonstrate that the difficulty of supervised learning with metric losses admits a tight finite characterization. In particular, we prove that the sample complexity of learning a hypothesis class can be detected by examining its finite projections. For realizable and agnostic learning with respect to a wide class of proper loss functions, we demonstrate an exact compactness result: a class is learnable with a given sample complexity precisely when the same is true of all its finite projections. For realizable learning with improper loss functions, we show that exact compactness of sample complexity can fail, and provide matching upper and lower bounds of a factor of 2 on the extent to which such sample complexities can differ. We conjecture that larger gaps are possible for the agnostic case. At the heart of our technical work is a compactness result concerning assignments of variables that maintain a class of functions below a target value, which generalizes Hall's classic matching theorem and may be of independent interest.

2402.10977

2024-02-15T18:20:42Z

Generative AI and Process Systems Engineering: The Next Frontier

This article explores how emerging generative artificial intelligence (GenAI) models, such as large language models (LLMs), can enhance solution methodologies within process systems engineering (PSE). These cutting-edge GenAI models, particularly foundation models (FMs), which are pre-trained on extensive, general-purpose datasets, offer versatile adaptability for a broad range of tasks, including responding to queries, image generation, and complex decision-making. Given the close relationship between advancements in PSE and developments in computing and systems technologies, exploring the synergy between GenAI and PSE is essential. We begin our discussion with a compact overview of both classic and emerging GenAI models, including FMs, and then dive into their applications within key PSE domains: synthesis and design, optimization and integration, and process monitoring and control. In each domain, we explore how GenAI models could potentially advance PSE methodologies, providing insights and prospects for each area. Furthermore, the article identifies and discusses potential challenges in fully leveraging GenAI within PSE, including multiscale modeling, data requirements, evaluation metrics and benchmarks, and trust and safety, thereby deepening the discourse on effective GenAI integration into systems analysis, design, optimization, operations, monitoring, and control. This paper provides a guide for future research focused on the applications of emerging GenAI in PSE.

2402.09698

2024-02-15T04:16:59Z

Combining Evidence Across Filtrations

In anytime-valid sequential inference, it is known that any admissible inference procedure must be based on test martingales and their composite generalization, called e-processes, which are nonnegative processes whose expectation at any arbitrary stopping time is upper-bounded by one. An e-process quantifies the accumulated evidence against a composite null hypothesis over a sequence of outcomes. This paper studies methods for combining e-processes that are computed using different information sets, i.e., filtrations, for a null hypothesis. Even though e-processes constructed on the same filtration can be combined effortlessly (e.g., by averaging), e-processes constructed on different filtrations cannot be combined as easily because their validity in a coarser filtration does not translate to validity in a finer filtration. We discuss three concrete examples of such e-processes in the literature: exchangeability tests, independence tests, and tests for evaluating and comparing forecasts with lags. Our main result establishes that these e-processes can be lifted into any finer filtration using adjusters, which are functions that allow betting on the running maximum of the accumulated wealth (thereby insuring against the loss of evidence). We also develop randomized adjusters that can improve the power of the resulting sequential inference procedure.

2402.10357

2024-02-15T22:59:14Z

Efficient Sampling on Riemannian Manifolds via Langevin MCMC

We study the task of efficiently sampling from a Gibbs distribution $d \pi^* = e^{-h} d {vol}_g$ over a Riemannian manifold $M$ via (geometric) Langevin MCMC; this algorithm involves computing exponential maps in random Gaussian directions and is efficiently implementable in practice. The key to our analysis of Langevin MCMC is a bound on the discretization error of the geometric Euler-Murayama scheme, assuming $\nabla h$ is Lipschitz and $M$ has bounded sectional curvature. Our error bound matches the error of Euclidean Euler-Murayama in terms of its stepsize dependence. Combined with a contraction guarantee for the geometric Langevin Diffusion under Kendall-Cranston coupling, we prove that the Langevin MCMC iterates lie within $\epsilon$-Wasserstein distance of $\pi^*$ after $\tilde{O}(\epsilon^{-2})$ steps, which matches the iteration complexity for Euclidean Langevin MCMC. Our results apply in general settings where $h$ can be nonconvex and $M$ can have negative Ricci curvature. Under additional assumptions that the Riemannian curvature tensor has bounded derivatives, and that $\pi^*$ satisfies a $CD(\cdot,\infty)$ condition, we analyze the stochastic gradient version of Langevin MCMC, and bound its iteration complexity by $\tilde{O}(\epsilon^{-2})$ as well.

2402.10435

2024-02-16T03:53:30Z

Dynamic Patch-aware Enrichment Transformer for Occluded Person Re-Identification

Person re-identification (re-ID) continues to pose a significant challenge, particularly in scenarios involving occlusions. Prior approaches aimed at tackling occlusions have predominantly focused on aligning physical body features through the utilization of external semantic cues. However, these methods tend to be intricate and susceptible to noise. To address the aforementioned challenges, we present an innovative end-to-end solution known as the Dynamic Patch-aware Enrichment Transformer (DPEFormer). This model effectively distinguishes human body information from occlusions automatically and dynamically, eliminating the need for external detectors or precise image alignment. Specifically, we introduce a dynamic patch token selection module (DPSM). DPSM utilizes a label-guided proxy token as an intermediary to identify informative occlusion-free tokens. These tokens are then selected for deriving subsequent local part features. To facilitate the seamless integration of global classification features with the finely detailed local features selected by DPSM, we introduce a novel feature blending module (FBM). FBM enhances feature representation through the complementary nature of information and the exploitation of part diversity. Furthermore, to ensure that DPSM and the entire DPEFormer can effectively learn with only identity labels, we also propose a Realistic Occlusion Augmentation (ROA) strategy. This strategy leverages the recent advances in the Segment Anything Model (SAM). As a result, it generates occlusion images that closely resemble real-world occlusions, greatly enhancing the subsequent contrastive learning process. Experiments on occluded and holistic re-ID benchmarks signify a substantial advancement of DPEFormer over existing state-of-the-art approaches. The code will be made publicly available.

2402.10454

2024-02-16T05:16:20Z

Optimizing Skin Lesion Classification via Multimodal Data and Auxiliary Task Integration

The rising global prevalence of skin conditions, some of which can escalate to life-threatening stages if not timely diagnosed and treated, presents a significant healthcare challenge. This issue is particularly acute in remote areas where limited access to healthcare often results in delayed treatment, allowing skin diseases to advance to more critical stages. One of the primary challenges in diagnosing skin diseases is their low inter-class variations, as many exhibit similar visual characteristics, making accurate classification challenging. This research introduces a novel multimodal method for classifying skin lesions, integrating smartphone-captured images with essential clinical and demographic information. This approach mimics the diagnostic process employed by medical professionals. A distinctive aspect of this method is the integration of an auxiliary task focused on super-resolution image prediction. This component plays a crucial role in refining visual details and enhancing feature extraction, leading to improved differentiation between classes and, consequently, elevating the overall effectiveness of the model. The experimental evaluations have been conducted using the PAD-UFES20 dataset, applying various deep-learning architectures. The results of these experiments not only demonstrate the effectiveness of the proposed method but also its potential applicability under-resourced healthcare environments.

2402.10476

2024-02-16T06:45:25Z

Spike-EVPR: Deep Spiking Residual Network with Cross-Representation Aggregation for Event-Based Visual Place Recognition

Event cameras have been successfully applied to visual place recognition (VPR) tasks by using deep artificial neural networks (ANNs) in recent years. However, previously proposed deep ANN architectures are often unable to harness the abundant temporal information presented in event streams. In contrast, deep spiking networks exhibit more intricate spatiotemporal dynamics and are inherently well-suited to process sparse asynchronous event streams. Unfortunately, directly inputting temporal-dense event volumes into the spiking network introduces excessive time steps, resulting in prohibitively high training costs for large-scale VPR tasks. To address the aforementioned issues, we propose a novel deep spiking network architecture called Spike-EVPR for event-based VPR tasks. First, we introduce two novel event representations tailored for SNN to fully exploit the spatio-temporal information from the event streams, and reduce the video memory occupation during training as much as possible. Then, to exploit the full potential of these two representations, we construct a Bifurcated Spike Residual Encoder (BSR-Encoder) with powerful representational capabilities to better extract the high-level features from the two event representations. Next, we introduce a Shared & Specific Descriptor Extractor (SSD-Extractor). This module is designed to extract features shared between the two representations and features specific to each. Finally, we propose a Cross-Descriptor Aggregation Module (CDA-Module) that fuses the above three features to generate a refined, robust global descriptor of the scene. Our experimental results indicate the superior performance of our Spike-EVPR compared to several existing EVPR pipelines on Brisbane-Event-VPR and DDD20 datasets, with the average Recall@1 increased by 7.61% on Brisbane and 13.20% on DDD20.

2402.10491

2024-02-16T07:48:35Z

Make a Cheap Scaling: A Self-Cascade Diffusion Model for Higher-Resolution Adaptation

Diffusion models have proven to be highly effective in image and video generation; however, they still face composition challenges when generating images of varying sizes due to single-scale training data. Adapting large pre-trained diffusion models for higher resolution demands substantial computational and optimization resources, yet achieving a generation capability comparable to low-resolution models remains elusive. This paper proposes a novel self-cascade diffusion model that leverages the rich knowledge gained from a well-trained low-resolution model for rapid adaptation to higher-resolution image and video generation, employing either tuning-free or cheap upsampler tuning paradigms. Integrating a sequence of multi-scale upsampler modules, the self-cascade diffusion model can efficiently adapt to a higher resolution, preserving the original composition and generation capabilities. We further propose a pivot-guided noise re-schedule strategy to speed up the inference process and improve local structural details. Compared to full fine-tuning, our approach achieves a 5X training speed-up and requires only an additional 0.002M tuning parameters. Extensive experiments demonstrate that our approach can quickly adapt to higher resolution image and video synthesis by fine-tuning for just 10k steps, with virtually no additional inference time.

2402.10520

2024-02-16T09:09:16Z

Real-Time Model-Based Quantitative Ultrasound and Radar

Ultrasound and radar signals are highly beneficial for medical imaging as they are non-invasive and non-ionizing. Traditional imaging techniques have limitations in terms of contrast and physical interpretation. Quantitative medical imaging can display various physical properties such as speed of sound, density, conductivity, and relative permittivity. This makes it useful for a wider range of applications, including improving cancer detection, diagnosing fatty liver, and fast stroke imaging. However, current quantitative imaging techniques that estimate physical properties from received signals, such as Full Waveform Inversion, are time-consuming and tend to converge to local minima, making them unsuitable for medical imaging. To address these challenges, we propose a neural network based on the physical model of wave propagation, which defines the relationship between the received signals and physical properties. Our network can reconstruct multiple physical properties in less than one second for complex and realistic scenarios, using data from only eight elements. We demonstrate the effectiveness of our approach for both radar and ultrasound signals.

2402.10534

2024-02-16T09:46:20Z

Using Left and Right Brains Together: Towards Vision and Language Planning

Large Language Models (LLMs) and Large Multi-modality Models (LMMs) have demonstrated remarkable decision masking capabilities on a variety of tasks. However, they inherently operate planning within the language space, lacking the vision and spatial imagination ability. In contrast, humans utilize both left and right hemispheres of the brain for language and visual planning during the thinking process. Therefore, we introduce a novel vision-language planning framework in this work to perform concurrent visual and language planning for tasks with inputs of any form. Our framework incorporates visual planning to capture intricate environmental details, while language planning enhances the logical coherence of the overall system. We evaluate the effectiveness of our framework across vision-language tasks, vision-only tasks, and language-only tasks. The results demonstrate the superior performance of our approach, indicating that the integration of visual and language planning yields better contextually aware task execution.

2402.10595

2024-02-16T11:28:50Z

Compact and De-biased Negative Instance Embedding for Multi-Instance Learning on Whole-Slide Image Classification

Whole-slide image (WSI) classification is a challenging task because 1) patches from WSI lack annotation, and 2) WSI possesses unnecessary variability, e.g., stain protocol. Recently, Multiple-Instance Learning (MIL) has made significant progress, allowing for classification based on slide-level, rather than patch-level, annotations. However, existing MIL methods ignore that all patches from normal slides are normal. Using this free annotation, we introduce a semi-supervision signal to de-bias the inter-slide variability and to capture the common factors of variation within normal patches. Because our method is orthogonal to the MIL algorithm, we evaluate our method on top of the recently proposed MIL algorithms and also compare the performance with other semi-supervised approaches. We evaluate our method on two public WSI datasets including Camelyon-16 and TCGA lung cancer and demonstrate that our approach significantly improves the predictive performance of existing MIL algorithms and outperforms other semi-supervised algorithms. We release our code at https://github.com/AITRICS/pathology_mil.

2402.10698

2024-02-16T13:59:07Z

Question-Instructed Visual Descriptions for Zero-Shot Video Question Answering

We present Q-ViD, a simple approach for video question answering (video QA), that unlike prior methods, which are based on complex architectures, computationally expensive pipelines or use closed models like GPTs, Q-ViD relies on a single instruction-aware open vision-language model (InstructBLIP) to tackle videoQA using frame descriptions. Specifically, we create captioning instruction prompts that rely on the target questions about the videos and leverage InstructBLIP to obtain video frame captions that are useful to the task at hand. Subsequently, we form descriptions of the whole video using the question-dependent frame captions, and feed that information, along with a question-answering prompt, to a large language model (LLM). The LLM is our reasoning module, and performs the final step of multiple-choice QA. Our simple Q-ViD framework achieves competitive or even higher performances than current state of the art models on a diverse range of videoQA benchmarks, including NExT-QA, STAR, How2QA, TVQA and IntentQA.

2402.10752

2024-02-16T15:19:39Z

STF: Spatio-Temporal Fusion Module for Improving Video Object Detection

Consecutive frames in a video contain redundancy, but they may also contain relevant complementary information for the detection task. The objective of our work is to leverage this complementary information to improve detection. Therefore, we propose a spatio-temporal fusion framework (STF). We first introduce multi-frame and single-frame attention modules that allow a neural network to share feature maps between nearby frames to obtain more robust object representations. Second, we introduce a dual-frame fusion module that merges feature maps in a learnable manner to improve them. Our evaluation is conducted on three different benchmarks including video sequences of moving road users. The performed experiments demonstrate that the proposed spatio-temporal fusion module leads to improved detection performance compared to baseline object detectors. Code is available at https://github.com/noreenanwar/STF-module

2402.10821

2024-02-16T16:47:21Z

Training Class-Imbalanced Diffusion Model Via Overlap Optimization

Diffusion models have made significant advances recently in high-quality image synthesis and related tasks. However, diffusion models trained on real-world datasets, which often follow long-tailed distributions, yield inferior fidelity for tail classes. Deep generative models, including diffusion models, are biased towards classes with abundant training images. To address the observed appearance overlap between synthesized images of rare classes and tail classes, we propose a method based on contrastive learning to minimize the overlap between distributions of synthetic images for different classes. We show variants of our probabilistic contrastive learning method can be applied to any class conditional diffusion model. We show significant improvement in image synthesis using our loss for multiple datasets with long-tailed distribution. Extensive experimental results demonstrate that the proposed method can effectively handle imbalanced data for diffusion-based generation and classification models. Our code and datasets will be publicly available at https://github.com/yanliang3612/DiffROP.

2402.10847

2024-02-16T17:36:56Z

Enhancement-Driven Pretraining for Robust Fingerprint Representation Learning

Fingerprint recognition stands as a pivotal component of biometric technology, with diverse applications from identity verification to advanced search tools. In this paper, we propose a unique method for deriving robust fingerprint representations by leveraging enhancement-based pre-training. Building on the achievements of U-Net-based fingerprint enhancement, our method employs a specialized encoder to derive representations from fingerprint images in a self-supervised manner. We further refine these representations, aiming to enhance the verification capabilities. Our experimental results, tested on publicly available fingerprint datasets, reveal a marked improvement in verification performance against established self-supervised training techniques. Our findings not only highlight the effectiveness of our method but also pave the way for potential advancements. Crucially, our research indicates that it is feasible to extract meaningful fingerprint representations from degraded images without relying on enhanced samples.

2402.10855

2024-02-16T17:51:13Z

Control Color: Multimodal Diffusion-based Interactive Image Colorization

Despite the existence of numerous colorization methods, several limitations still exist, such as lack of user interaction, inflexibility in local colorization, unnatural color rendering, insufficient color variation, and color overflow. To solve these issues, we introduce Control Color (CtrlColor), a multi-modal colorization method that leverages the pre-trained Stable Diffusion (SD) model, offering promising capabilities in highly controllable interactive image colorization. While several diffusion-based methods have been proposed, supporting colorization in multiple modalities remains non-trivial. In this study, we aim to tackle both unconditional and conditional image colorization (text prompts, strokes, exemplars) and address color overflow and incorrect color within a unified framework. Specifically, we present an effective way to encode user strokes to enable precise local color manipulation and employ a practical way to constrain the color distribution similar to exemplars. Apart from accepting text prompts as conditions, these designs add versatility to our approach. We also introduce a novel module based on self-attention and a content-guided deformable autoencoder to address the long-standing issues of color overflow and inaccurate coloring. Extensive comparisons show that our model outperforms state-of-the-art image colorization methods both qualitatively and quantitatively.

2402.10896

2024-02-16T18:54:47Z

PaLM2-VAdapter: Progressively Aligned Language Model Makes a Strong Vision-language Adapter

This paper demonstrates that a progressively aligned language model can effectively bridge frozen vision encoders and large language models (LLMs). While the fundamental architecture and pre-training methods of vision encoders and LLMs have been extensively studied, the architecture and training strategy of vision-language adapters vary significantly across recent works. Our research undertakes a thorough exploration of the state-of-the-art perceiver resampler architecture and builds a strong baseline. However, we observe that the vision-language alignment with perceiver resampler exhibits slow convergence and limited scalability with a lack of direct supervision. To address this issue, we propose PaLM2-VAdapter, employing a progressively aligned language model as the vision-language adapter. Compared to the strong baseline with perceiver resampler, our method empirically shows faster convergence, higher performance, and stronger scalability. Extensive experiments across various Visual Question Answering (VQA) and captioning tasks on both images and videos demonstrate that our model exhibits state-of-the-art visual understanding and multi-modal reasoning capabilities. Notably, our method achieves these advancements with 30~70% fewer parameters than the state-of-the-art large vision-language models, marking a significant efficiency improvement.

2402.11083

2024-02-16T21:17:42Z

VQAttack: Transferable Adversarial Attacks on Visual Question Answering via Pre-trained Models

Visual Question Answering (VQA) is a fundamental task in computer vision and natural language process fields. Although the ``pre-training & finetuning'' learning paradigm significantly improves the VQA performance, the adversarial robustness of such a learning paradigm has not been explored. In this paper, we delve into a new problem: using a pre-trained multimodal source model to create adversarial image-text pairs and then transferring them to attack the target VQA models. Correspondingly, we propose a novel VQAttack model, which can iteratively generate both image and text perturbations with the designed modules: the large language model (LLM)-enhanced image attack and the cross-modal joint attack module. At each iteration, the LLM-enhanced image attack module first optimizes the latent representation-based loss to generate feature-level image perturbations. Then it incorporates an LLM to further enhance the image perturbations by optimizing the designed masked answer anti-recovery loss. The cross-modal joint attack module will be triggered at a specific iteration, which updates the image and text perturbations sequentially. Notably, the text perturbation updates are based on both the learned gradients in the word embedding space and word synonym-based substitution. Experimental results on two VQA datasets with five validated models demonstrate the effectiveness of the proposed VQAttack in the transferable attack setting, compared with state-of-the-art baselines. This work reveals a significant blind spot in the ``pre-training & fine-tuning'' paradigm on VQA tasks. Source codes will be released.

2402.11095

2024-02-16T21:48:17Z

GIM: Learning Generalizable Image Matcher From Internet Videos

Image matching is a fundamental computer vision problem. While learning-based methods achieve state-of-the-art performance on existing benchmarks, they generalize poorly to in-the-wild images. Such methods typically need to train separate models for different scene types and are impractical when the scene type is unknown in advance. One of the underlying problems is the limited scalability of existing data construction pipelines, which limits the diversity of standard image matching datasets. To address this problem, we propose GIM, a self-training framework for learning a single generalizable model based on any image matching architecture using internet videos, an abundant and diverse data source. Given an architecture, GIM first trains it on standard domain-specific datasets and then combines it with complementary matching methods to create dense labels on nearby frames of novel videos. These labels are filtered by robust fitting, and then enhanced by propagating them to distant frames. The final model is trained on propagated data with strong augmentations. We also propose ZEB, the first zero-shot evaluation benchmark for image matching. By mixing data from diverse domains, ZEB can thoroughly assess the cross-domain generalization performance of different methods. Applying GIM consistently improves the zero-shot performance of 3 state-of-the-art image matching architectures; with 50 hours of YouTube videos, the relative zero-shot performance improves by 8.4%-18.1%. GIM also enables generalization to extreme cross-domain data such as Bird Eye View (BEV) images of projected 3D point clouds (Fig. 1(c)). More importantly, our single zero-shot model consistently outperforms domain-specific baselines when evaluated on downstream tasks inherent to their respective domains. The video presentation is available at https://www.youtube.com/watch?v=FU_MJLD8LeY.

2402.10376

2024-02-16T00:04:36Z

Interpreting CLIP with Sparse Linear Concept Embeddings (SpLiCE)

CLIP embeddings have demonstrated remarkable performance across a wide range of computer vision tasks. However, these high-dimensional, dense vector representations are not easily interpretable, restricting their usefulness in downstream applications that require transparency. In this work, we empirically show that CLIP's latent space is highly structured, and consequently that CLIP representations can be decomposed into their underlying semantic components. We leverage this understanding to propose a novel method, Sparse Linear Concept Embeddings (SpLiCE), for transforming CLIP representations into sparse linear combinations of human-interpretable concepts. Distinct from previous work, SpLiCE does not require concept labels and can be applied post hoc. Through extensive experimentation with multiple real-world datasets, we validate that the representations output by SpLiCE can explain and even replace traditional dense CLIP representations, maintaining equivalent downstream performance while significantly improving their interpretability. We also demonstrate several use cases of SpLiCE representations including detecting spurious correlations, model editing, and quantifying semantic shifts in datasets.

2402.10404

2024-02-16T02:12:20Z

Explaining generative diffusion models via visual analysis for interpretable decision-making process

Diffusion models have demonstrated remarkable performance in generation tasks. Nevertheless, explaining the diffusion process remains challenging due to it being a sequence of denoising noisy images that are difficult for experts to interpret. To address this issue, we propose the three research questions to interpret the diffusion process from the perspective of the visual concepts generated by the model and the region where the model attends in each time step. We devise tools for visualizing the diffusion process and answering the aforementioned research questions to render the diffusion process human-understandable. We show how the output is progressively generated in the diffusion process by explaining the level of denoising and highlighting relationships to foundational visual concepts at each time step through the results of experiments with various visual analyses using the tools. Throughout the training of the diffusion model, the model learns diverse visual concepts corresponding to each time-step, enabling the model to predict varying levels of visual concepts at different stages. We substantiate our tools using Area Under Cover (AUC) score, correlation quantification, and cross-attention mapping. Our findings provide insights into the diffusion process and pave the way for further research into explainable diffusion mechanisms.

2402.10478

2024-02-16T06:57:03Z

CodaMal: Contrastive Domain Adaptation for Malaria Detection in Low-Cost Microscopes

Malaria is a major health issue worldwide, and its diagnosis requires scalable solutions that can work effectively with low-cost microscopes (LCM). Deep learning-based methods have shown success in computer-aided diagnosis from microscopic images. However, these methods need annotated images that show cells affected by malaria parasites and their life stages. Annotating images from LCM significantly increases the burden on medical experts compared to annotating images from high-cost microscopes (HCM). For this reason, a practical solution would be trained on HCM images which should generalize well on LCM images during testing. While earlier methods adopted a multi-stage learning process, they did not offer an end-to-end approach. In this work, we present an end-to-end learning framework, named CodaMal (Contrastive Domain Adpation for Malaria). In order to bridge the gap between HCM (training) and LCM (testing), we propose a domain adaptive contrastive loss. It reduces the domain shift by promoting similarity between the representations of HCM and its corresponding LCM image, without imposing an additional annotation burden. In addition, the training objective includes object detection objectives with carefully designed augmentations, ensuring the accurate detection of malaria parasites. On the publicly available large-scale M5-dataset, our proposed method shows a significant improvement of 16% over the state-of-the-art methods in terms of the mean average precision metric (mAP), provides 21x speed up during inference, and requires only half learnable parameters than the prior methods. Our code is publicly available.

2402.10483

2024-02-16T07:13:24Z

GaussianHair: Hair Modeling and Rendering with Light-aware Gaussians

Hairstyle reflects culture and ethnicity at first glance. In the digital era, various realistic human hairstyles are also critical to high-fidelity digital human assets for beauty and inclusivity. Yet, realistic hair modeling and real-time rendering for animation is a formidable challenge due to its sheer number of strands, complicated structures of geometry, and sophisticated interaction with light. This paper presents GaussianHair, a novel explicit hair representation. It enables comprehensive modeling of hair geometry and appearance from images, fostering innovative illumination effects and dynamic animation capabilities. At the heart of GaussianHair is the novel concept of representing each hair strand as a sequence of connected cylindrical 3D Gaussian primitives. This approach not only retains the hair's geometric structure and appearance but also allows for efficient rasterization onto a 2D image plane, facilitating differentiable volumetric rendering. We further enhance this model with the "GaussianHair Scattering Model", adept at recreating the slender structure of hair strands and accurately capturing their local diffuse color in uniform lighting. Through extensive experiments, we substantiate that GaussianHair achieves breakthroughs in both geometric and appearance fidelity, transcending the limitations encountered in state-of-the-art methods for hair reconstruction. Beyond representation, GaussianHair extends to support editing, relighting, and dynamic rendering of hair, offering seamless integration with conventional CG pipeline workflows. Complementing these advancements, we have compiled an extensive dataset of real human hair, each with meticulously detailed strand geometry, to propel further research in this field.

2402.10665

2024-02-16T13:14:12Z

Selective Prediction for Semantic Segmentation using Post-Hoc Confidence Estimation and Its Performance under Distribution Shift

Semantic segmentation plays a crucial role in various computer vision applications, yet its efficacy is often hindered by the lack of high-quality labeled data. To address this challenge, a common strategy is to leverage models trained on data from different populations, such as publicly available datasets. This approach, however, leads to the distribution shift problem, presenting a reduced performance on the population of interest. In scenarios where model errors can have significant consequences, selective prediction methods offer a means to mitigate risks and reduce reliance on expert supervision. This paper investigates selective prediction for semantic segmentation in low-resource settings, thus focusing on post-hoc confidence estimators applied to pre-trained models operating under distribution shift. We propose a novel image-level confidence measure tailored for semantic segmentation and demonstrate its effectiveness through experiments on three medical imaging tasks. Our findings show that post-hoc confidence estimators offer a cost-effective approach to reducing the impacts of distribution shift.

2402.10717

2024-02-16T14:19:33Z

BioFusionNet: Deep Learning-Based Survival Risk Stratification in ER+ Breast Cancer Through Multifeature and Multimodal Data Fusion

Breast cancer is a significant health concern affecting millions of women worldwide. Accurate survival risk stratification plays a crucial role in guiding personalised treatment decisions and improving patient outcomes. Here we present BioFusionNet, a deep learning framework that fuses image-derived features with genetic and clinical data to achieve a holistic patient profile and perform survival risk stratification of ER+ breast cancer patients. We employ multiple self-supervised feature extractors, namely DINO and MoCoV3, pretrained on histopathology patches to capture detailed histopathological image features. We then utilise a variational autoencoder (VAE) to fuse these features, and harness the latent space of the VAE to feed into a self-attention network, generating patient-level features. Next, we develop a co-dual-cross-attention mechanism to combine the histopathological features with genetic data, enabling the model to capture the interplay between them. Additionally, clinical data is incorporated using a feed-forward network (FFN), further enhancing predictive performance and achieving comprehensive multimodal feature integration. Furthermore, we introduce a weighted Cox loss function, specifically designed to handle imbalanced survival data, which is a common challenge in the field. The proposed model achieves a mean concordance index (C-index) of 0.77 and a time-dependent area under the curve (AUC) of 0.84, outperforming state-of-the-art methods. It predicts risk (high versus low) with prognostic significance for overall survival (OS) in univariate analysis (HR=2.99, 95% CI: 1.88--4.78, p<0.005), and maintains independent significance in multivariate analysis incorporating standard clinicopathological variables (HR=2.91, 95% CI: 1.80--4.68, p<0.005). The proposed method not only improves model performance but also addresses a critical gap in handling imbalanced data.

2402.10728

2024-02-16T14:44:40Z

Semi-weakly-supervised neural network training for medical image registration

For training registration networks, weak supervision from segmented corresponding regions-of-interest (ROIs) have been proven effective for (a) supplementing unsupervised methods, and (b) being used independently in registration tasks in which unsupervised losses are unavailable or ineffective. This correspondence-informing supervision entails cost in annotation that requires significant specialised effort. This paper describes a semi-weakly-supervised registration pipeline that improves the model performance, when only a small corresponding-ROI-labelled dataset is available, by exploiting unlabelled image pairs. We examine two types of augmentation methods by perturbation on network weights and image resampling, such that consistency-based unsupervised losses can be applied on unlabelled data. The novel WarpDDF and RegCut approaches are proposed to allow commutative perturbation between an image pair and the predicted spatial transformation (i.e. respective input and output of registration networks), distinct from existing perturbation methods for classification or segmentation. Experiments using 589 male pelvic MR images, labelled with eight anatomical ROIs, show the improvement in registration performance and the ablated contributions from the individual strategies. Furthermore, this study attempts to construct one of the first computational atlases for pelvic structures, enabled by registering inter-subject MRs, and quantifies the significant differences due to the proposed semi-weak supervision with a discussion on the potential clinical use of example atlas-derived statistics.

2402.10776

2024-02-16T15:58:45Z

In-Vivo Hyperspectral Human Brain Image Database for Brain Cancer Detection

The use of hyperspectral imaging for medical applications is becoming more common in recent years. One of the main obstacles that researchers find when developing hyperspectral algorithms for medical applications is the lack of specific, publicly available, and hyperspectral medical data. The work described in this paper was developed within the framework of the European project HELICoiD (HypErspectraL Imaging Cancer Detection), which had as a main goal the application of hyperspectral imaging to the delineation of brain tumors in real-time during neurosurgical operations. In this paper, the methodology followed to generate the first hyperspectral database of in-vivo human brain tissues is presented. Data was acquired employing a customized hyperspectral acquisition system capable of capturing information in the Visual and Near InfraRed (VNIR) range from 400 to 1000 nm. Repeatability was assessed for the cases where two images of the same scene were captured consecutively. The analysis reveals that the system works more efficiently in the spectral range between 450 and 900 nm. A total of 36 hyperspectral images from 22 different patients were obtained. From these data, more than 300 000 spectral signatures were labeled employing a semi-automatic methodology based on the spectral angle mapper algorithm. Four different classes were defined: normal tissue, tumor tissue, blood vessel, and background elements. All the hyperspectral data has been made available in a public repository.

2402.10798

2024-02-16T16:21:15Z

VATr++: Choose Your Words Wisely for Handwritten Text Generation

Styled Handwritten Text Generation (HTG) has received significant attention in recent years, propelled by the success of learning-based solutions employing GANs, Transformers, and, preliminarily, Diffusion Models. Despite this surge in interest, there remains a critical yet understudied aspect - the impact of the input, both visual and textual, on the HTG model training and its subsequent influence on performance. This study delves deeper into a cutting-edge Styled-HTG approach, proposing strategies for input preparation and training regularization that allow the model to achieve better performance and generalize better. These aspects are validated through extensive analysis on several different settings and datasets. Moreover, in this work, we go beyond performance optimization and address a significant hurdle in HTG research - the lack of a standardized evaluation protocol. In particular, we propose a standardization of the evaluation protocol for HTG and conduct a comprehensive benchmarking of existing approaches. By doing so, we aim to establish a foundation for fair and meaningful comparisons between HTG strategies, fostering progress in the field.

2402.10814

2024-02-16T16:37:48Z

Associative Memories in the Feature Space

An autoassociative memory model is a function that, given a set of data points, takes as input an arbitrary vector and outputs the most similar data point from the memorized set. However, popular memory models fail to retrieve images even when the corruption is mild and easy to detect for a human evaluator. This is because similarities are evaluated in the raw pixel space, which does not contain any semantic information about the images. This problem can be easily solved by computing \emph{similarities} in an embedding space instead of the pixel space. We show that an effective way of computing such embeddings is via a network pretrained with a contrastive loss. As the dimension of embedding spaces is often significantly smaller than the pixel space, we also have a faster computation of similarity scores. We test this method on complex datasets such as CIFAR10 and STL10. An additional drawback of current models is the need of storing the whole dataset in the pixel space, which is often extremely large. We relax this condition and propose a class of memory models that only stores low-dimensional semantic embeddings, and uses them to retrieve similar, but not identical, memories. We demonstrate a proof of concept of this method on a simple task on the MNIST dataset.

2402.10865

2024-02-16T18:01:43Z

Multi-Model 3D Registration: Finding Multiple Moving Objects in Cluttered Point Clouds

We investigate a variation of the 3D registration problem, named multi-model 3D registration. In the multi-model registration problem, we are given two point clouds picturing a set of objects at different poses (and possibly including points belonging to the background) and we want to simultaneously reconstruct how all objects moved between the two point clouds. This setup generalizes standard 3D registration where one wants to reconstruct a single pose, e.g., the motion of the sensor picturing a static scene. Moreover, it provides a mathematically grounded formulation for relevant robotics applications, e.g., where a depth sensor onboard a robot perceives a dynamic scene and has the goal of estimating its own motion (from the static portion of the scene) while simultaneously recovering the motion of all dynamic objects. We assume a correspondence-based setup where we have putative matches between the two point clouds and consider the practical case where these correspondences are plagued with outliers. We then propose a simple approach based on Expectation-Maximization (EM) and establish theoretical conditions under which the EM approach converges to the ground truth. We evaluate the approach in simulated and real datasets ranging from table-top scenes to self-driving scenarios and demonstrate its effectiveness when combined with state-of-the-art scene flow methods to establish dense correspondences.

2402.10882

2024-02-16T18:36:36Z

Universal Prompt Optimizer for Safe Text-to-Image Generation

Text-to-Image (T2I) models have shown great performance in generating images based on textual prompts. However, these models are vulnerable to unsafe input to generate unsafe content like sexual, harassment and illegal-activity images. Existing studies based on image checker, model fine-tuning and embedding blocking are impractical in real-world applications. Hence, \textit{we propose the first universal prompt optimizer for safe T2I generation in black-box scenario}. We first construct a dataset consisting of toxic-clean prompt pairs by GPT-3.5 Turbo. To guide the optimizer to have the ability of converting toxic prompt to clean prompt while preserving semantic information, we design a novel reward function measuring toxicity and text alignment of generated images and train the optimizer through Proximal Policy Optimization. Experiments show that our approach can effectively reduce the likelihood of various T2I models in generating inappropriate images, with no significant impact on text alignment. It is also flexible to be combined with methods to achieve better performance.

2402.10887

2024-02-16T18:43:39Z

Weak-Mamba-UNet: Visual Mamba Makes CNN and ViT Work Better for Scribble-based Medical Image Segmentation

Medical image segmentation is increasingly reliant on deep learning techniques, yet the promising performance often come with high annotation costs. This paper introduces Weak-Mamba-UNet, an innovative weakly-supervised learning (WSL) framework that leverages the capabilities of Convolutional Neural Network (CNN), Vision Transformer (ViT), and the cutting-edge Visual Mamba (VMamba) architecture for medical image segmentation, especially when dealing with scribble-based annotations. The proposed WSL strategy incorporates three distinct architecture but same symmetrical encoder-decoder networks: a CNN-based UNet for detailed local feature extraction, a Swin Transformer-based SwinUNet for comprehensive global context understanding, and a VMamba-based Mamba-UNet for efficient long-range dependency modeling. The key concept of this framework is a collaborative and cross-supervisory mechanism that employs pseudo labels to facilitate iterative learning and refinement across the networks. The effectiveness of Weak-Mamba-UNet is validated on a publicly available MRI cardiac segmentation dataset with processed scribble annotations, where it surpasses the performance of a similar WSL framework utilizing only UNet or SwinUNet. This highlights its potential in scenarios with sparse or imprecise annotations. The source code is made publicly accessible.

2402.10894

2024-02-16T18:51:42Z

Fusion of Diffusion Weighted MRI and Clinical Data for Predicting Functional Outcome after Acute Ischemic Stroke with Deep Contrastive Learning

Stroke is a common disabling neurological condition that affects about one-quarter of the adult population over age 25; more than half of patients still have poor outcomes, such as permanent functional dependence or even death, after the onset of acute stroke. The aim of this study is to investigate the efficacy of diffusion-weighted MRI modalities combining with structured health profile on predicting the functional outcome to facilitate early intervention. A deep fusion learning network is proposed with two-stage training: the first stage focuses on cross-modality representation learning and the second stage on classification. Supervised contrastive learning is exploited to learn discriminative features that separate the two classes of patients from embeddings of individual modalities and from the fused multimodal embedding. The network takes as the input DWI and ADC images, and structured health profile data. The outcome is the prediction of the patient needing long-term care at 3 months after the onset of stroke. Trained and evaluated with a dataset of 3297 patients, our proposed fusion model achieves 0.87, 0.80 and 80.45% for AUC, F1-score and accuracy, respectively, outperforming existing models that consolidate both imaging and structured data in the medical domain. If trained with comprehensive clinical variables, including NIHSS and comorbidities, the gain from images on making accurate prediction is not considered substantial, but significant. However, diffusion-weighted MRI can replace NIHSS to achieve comparable level of accuracy combining with other readily available clinical variables for better generalization.

2402.11036

2024-02-16T19:29:43Z

Occlusion Resilient 3D Human Pose Estimation

Occlusions remain one of the key challenges in 3D body pose estimation from single-camera video sequences. Temporal consistency has been extensively used to mitigate their impact but the existing algorithms in the literature do not explicitly model them. Here, we apply this by representing the deforming body as a spatio-temporal graph. We then introduce a refinement network that performs graph convolutions over this graph to output 3D poses. To ensure robustness to occlusions, we train this network with a set of binary masks that we use to disable some of the edges as in drop-out techniques. In effect, we simulate the fact that some joints can be hidden for periods of time and train the network to be immune to that. We demonstrate the effectiveness of this approach compared to state-of-the-art techniques that infer poses from single-camera sequences.

2402.11058

2024-02-16T20:14:47Z

II-MMR: Identifying and Improving Multi-modal Multi-hop Reasoning in Visual Question Answering

Visual Question Answering (VQA) often involves diverse reasoning scenarios across Vision and Language (V&L). Most prior VQA studies, however, have merely focused on assessing the model's overall accuracy without evaluating it on different reasoning cases. Furthermore, some recent works observe that conventional Chain-of-Thought (CoT) prompting fails to generate effective reasoning for VQA, especially for complex scenarios requiring multi-hop reasoning. In this paper, we propose II-MMR, a novel idea to identify and improve multi-modal multi-hop reasoning in VQA. In specific, II-MMR takes a VQA question with an image and finds a reasoning path to reach its answer using two novel language promptings: (i) answer prediction-guided CoT prompt, or (ii) knowledge triplet-guided prompt. II-MMR then analyzes this path to identify different reasoning cases in current VQA benchmarks by estimating how many hops and what types (i.e., visual or beyond-visual) of reasoning are required to answer the question. On popular benchmarks including GQA and A-OKVQA, II-MMR observes that most of their VQA questions are easy to answer, simply demanding "single-hop" reasoning, whereas only a few questions require "multi-hop" reasoning. Moreover, while the recent V&L model struggles with such complex multi-hop reasoning questions even using the traditional CoT method, II-MMR shows its effectiveness across all reasoning cases in both zero-shot and fine-tuning settings.

2402.11093

2024-02-16T21:39:28Z

Modular Graph Extraction for Handwritten Circuit Diagram Images

As digitization in engineering progressed, circuit diagrams (also referred to as schematics) are typically developed and maintained in computer-aided engineering (CAE) systems, thus allowing for automated verification, simulation and further processing in downstream engineering steps. However, apart from printed legacy schematics, hand-drawn circuit diagrams are still used today in the educational domain, where they serve as an easily accessible mean for trainees and students to learn drawing this type of diagrams. Furthermore, hand-drawn schematics are typically used in examinations due to legal constraints. In order to harness the capabilities of digital circuit representations, automated means for extracting the electrical graph from raster graphics are required. While respective approaches have been proposed in literature, they are typically conducted on small or non-disclosed datasets. This paper describes a modular end-to-end solution on a larger, public dataset, in which approaches for the individual sub-tasks are evaluated to form a new baseline. These sub-tasks include object detection (for electrical symbols and texts), binary segmentation (drafter's stroke vs. background), handwritten character recognition and orientation regression for electrical symbols and texts. Furthermore, computer-vision graph assembly and rectification algorithms are presented. All methods are integrated in a publicly available prototype.

2402.10403

2024-02-16T02:01:24Z

Polyhedral Complex Derivation from Piecewise Trilinear Networks

Recent advancements in visualizing deep neural networks provide insights into their structures and mesh extraction from Continuous Piecewise Affine (CPWA) functions. Meanwhile, developments in neural surface representation learning incorporate non-linear positional encoding, addressing issues like spectral bias; however, this poses challenges in applying mesh extraction techniques based on CPWA functions. Focusing on trilinear interpolating methods as positional encoding, we present theoretical insights and an analytical mesh extraction, showing the transformation of hypersurfaces to flat planes within the trilinear region under the eikonal constraint. Moreover, we introduce a method for approximating intersecting points among three hypersurfaces contributing to broader applications. We empirically validate correctness and parsimony through chamfer distance and efficiency, and angular distance, while examining the correlation between the eikonal loss and the planarity of the hypersurfaces.

2402.10425

2024-02-16T03:22:58Z

DABS-LS: Deep Atlas-Based Segmentation Using Regional Level Set Self-Supervision

Cochlear implants (CIs) are neural prosthetics used to treat patients with severe-to-profound hearing loss. Patient-specific modeling of CI stimulation of the auditory nerve fiber (ANFs) can help audiologists improve the CI programming. These models require localization of the ANFs relative to surrounding anatomy and the CI. Localization is challenging because the ANFs are so small they are not directly visible in clinical imaging. In this work, we hypothesize the position of the ANFs can be accurately inferred from the location of the internal auditory canal (IAC), which has high contrast in CT, since the ANFs pass through this canal between the cochlea and the brain. Inspired by VoxelMorph, in this paper we propose a deep atlas-based IAC segmentation network. We create a single atlas in which the IAC and ANFs are pre-localized. Our network is trained to produce deformation fields (DFs) mapping coordinates from the atlas to new target volumes and that accurately segment the IAC. We hypothesize that DFs that accurately segment the IAC in target images will also facilitate accurate atlas-based localization of the ANFs. As opposed to VoxelMorph, which aims to produce DFs that accurately register the entire volume, our novel contribution is an entirely self-supervised training scheme that aims to produce DFs that accurately segment the target structure. This self-supervision is facilitated using a regional level set (LS) inspired loss function. We call our method Deep Atlas Based Segmentation using Level Sets (DABS-LS). Results show that DABS-LS outperforms VoxelMorph for IAC segmentation. Tests with publicly available datasets for trachea and kidney segmentation also show significant improvement in segmentation accuracy, demonstrating the generalizability of the method.

2402.10470

2024-02-16T06:22:44Z

Theoretical Understanding of Learning from Adversarial Perturbations

It is not fully understood why adversarial examples can deceive neural networks and transfer between different networks. To elucidate this, several studies have hypothesized that adversarial perturbations, while appearing as noises, contain class features. This is supported by empirical evidence showing that networks trained on mislabeled adversarial examples can still generalize well to correctly labeled test samples. However, a theoretical understanding of how perturbations include class features and contribute to generalization is limited. In this study, we provide a theoretical framework for understanding learning from perturbations using a one-hidden-layer network trained on mutually orthogonal samples. Our results highlight that various adversarial perturbations, even perturbations of a few pixels, contain sufficient class features for generalization. Moreover, we reveal that the decision boundary when learning from perturbations matches that from standard samples except for specific regions under mild conditions. The code is available at https://github.com/s-kumano/learning-from-adversarial-perturbations.

2402.10553

2024-02-16T10:35:01Z

A novel integrated industrial approach with cobots in the age of industry 4.0 through conversational interaction and computer vision

From robots that replace workers to robots that serve as helpful colleagues, the field of robotic automation is experiencing a new trend that represents a huge challenge for component manufacturers. The contribution starts from an innovative vision that sees an ever closer collaboration between Cobot, able to do a specific physical job with precision, the AI world, able to analyze information and support the decision-making process, and the man able to have a strategic vision of the future.

2402.10580

2024-02-16T11:09:16Z

Efficient Multi-task Uncertainties for Joint Semantic Segmentation and Monocular Depth Estimation

Quantifying the predictive uncertainty emerged as a possible solution to common challenges like overconfidence or lack of explainability and robustness of deep neural networks, albeit one that is often computationally expensive. Many real-world applications are multi-modal in nature and hence benefit from multi-task learning. In autonomous driving, for example, the joint solution of semantic segmentation and monocular depth estimation has proven to be valuable. In this work, we first combine different uncertainty quantification methods with joint semantic segmentation and monocular depth estimation and evaluate how they perform in comparison to each other. Additionally, we reveal the benefits of multi-task learning with regard to the uncertainty quality compared to solving both tasks separately. Based on these insights, we introduce EMUFormer, a novel student-teacher distillation approach for joint semantic segmentation and monocular depth estimation as well as efficient multi-task uncertainty quantification. By implicitly leveraging the predictive uncertainties of the teacher, EMUFormer achieves new state-of-the-art results on Cityscapes and NYUv2 and additionally estimates high-quality predictive uncertainties for both tasks that are comparable or superior to a Deep Ensemble despite being an order of magnitude more efficient.

2402.10609

2024-02-16T11:54:34Z

U$^2$MRPD: Unsupervised undersampled MRI reconstruction by prompting a large latent diffusion model

Implicit visual knowledge in a large latent diffusion model (LLDM) pre-trained on natural images is rich and hypothetically universal to natural and medical images. To test this hypothesis, we introduce a novel framework for Unsupervised Undersampled MRI Reconstruction by Prompting a pre-trained large latent Diffusion model ( U$^2$MRPD). Existing data-driven, supervised undersampled MRI reconstruction networks are typically of limited generalizability and adaptability toward diverse data acquisition scenarios; yet U$^2$MRPD supports image-specific MRI reconstruction by prompting an LLDM with an MRSampler tailored for complex-valued MRI images. With any single-source or diverse-source MRI dataset, U$^2$MRPD's performance is further boosted by an MRAdapter while keeping the generative image priors intact. Experiments on multiple datasets show that U$^2$MRPD achieves comparable or better performance than supervised and MRI diffusion methods on in-domain datasets while demonstrating the best generalizability on out-of-domain datasets. To the best of our knowledge, U$^2$MRPD is the {\bf first} unsupervised method that demonstrates the universal prowess of a LLDM, %trained on magnitude-only natural images in medical imaging, attaining the best adaptability for both MRI database-free and database-available scenarios and generalizability towards out-of-domain data.

2402.10747

2024-02-16T15:13:30Z

Fully Differentiable Lagrangian Convolutional Neural Network for Continuity-Consistent Physics-Informed Precipitation Nowcasting

This paper presents a convolutional neural network model for precipitation nowcasting that combines data-driven learning with physics-informed domain knowledge. We propose LUPIN, a Lagrangian Double U-Net for Physics-Informed Nowcasting, that draws from existing extrapolation-based nowcasting methods and implements the Lagrangian coordinate system transformation of the data in a fully differentiable and GPU-accelerated manner to allow for real-time end-to-end training and inference. Based on our evaluation, LUPIN matches and exceeds the performance of the chosen benchmark, opening the door for other Lagrangian machine learning models.

2402.10851

2024-02-16T17:44:11Z

HistoSegCap: Capsules for Weakly-Supervised Semantic Segmentation of Histological Tissue Type in Whole Slide Images

Digital pathology involves converting physical tissue slides into high-resolution Whole Slide Images (WSIs), which pathologists analyze for disease-affected tissues. However, large histology slides with numerous microscopic fields pose challenges for visual search. To aid pathologists, Computer Aided Diagnosis (CAD) systems offer visual assistance in efficiently examining WSIs and identifying diagnostically relevant regions. This paper presents a novel histopathological image analysis method employing Weakly Supervised Semantic Segmentation (WSSS) based on Capsule Networks, the first such application. The proposed model is evaluated using the Atlas of Digital Pathology (ADP) dataset and its performance is compared with other histopathological semantic segmentation methodologies. The findings underscore the potential of Capsule Networks in enhancing the precision and efficiency of histopathological image analysis. Experimental results show that the proposed model outperforms traditional methods in terms of accuracy and the mean Intersection-over-Union (mIoU) metric.

2402.10884

2024-02-16T18:42:08Z

Multi-modal preference alignment remedies regression of visual instruction tuning on language model

In production, multi-modal large language models (MLLMs) are expected to support multi-turn queries of interchanging image and text modalities. However, the current MLLMs trained with visual-question-answering (VQA) datasets could suffer from degradation, as VQA datasets lack the diversity and complexity of the original text instruction datasets which the underlying language model had been trained with. To address this challenging degradation, we first collect a lightweight (6k entries) VQA preference dataset where answers were annotated by Gemini for 5 quality metrics in a granular fashion, and investigate standard Supervised Fine-tuning, rejection sampling, Direct Preference Optimization (DPO), and SteerLM. Our findings indicate that the with DPO we are able to surpass instruction-following capabilities of the language model, achieving a 6.73 score on MT-Bench, compared to Vicuna's 6.57 and LLaVA's 5.99 despite small data scale. This enhancement in textual instruction proficiency correlates with boosted visual instruction performance (+4.9\% on MM-Vet, +6\% on LLaVA-Bench), with minimal alignment tax on visual knowledge benchmarks compared to previous RLHF approach. In conclusion, we propose a distillation-based multi-modal alignment model with fine-grained annotations on a small dataset that reconciles the textual and visual performance of MLLMs, restoring and boosting language capability after visual instruction tuning.

2402.11089

2024-02-16T21:32:27Z

The Male CEO and the Female Assistant: Probing Gender Biases in Text-To-Image Models Through Paired Stereotype Test

Recent large-scale Text-To-Image (T2I) models such as DALLE-3 demonstrate great potential in new applications, but also face unprecedented fairness challenges. Prior studies revealed gender biases in single-person image generation, but T2I model applications might require portraying two or more people simultaneously. Potential biases in this setting remain unexplored, leading to fairness-related risks in usage. To study these underlying facets of gender biases in T2I models, we propose a novel Paired Stereotype Test (PST) bias evaluation framework. PST prompts the model to generate two individuals in the same image. They are described with two social identities that are stereotypically associated with the opposite gender. Biases can then be measured by the level of conformation to gender stereotypes in generated images. Using PST, we evaluate DALLE-3 from 2 perspectives: biases in gendered occupation and biases in organizational power. Despite seemingly fair or even anti-stereotype single-person generations, PST still unveils gendered occupational and power associations. Moreover, compared to single-person settings, DALLE-3 generates noticeably more masculine figures under PST for individuals with male-stereotypical identities. PST is therefore effective in revealing underlying gender biases in DALLE-3 that single-person settings cannot capture. Our findings reveal the complicated patterns of gender biases in modern T2I models, further highlighting the critical fairness challenges in multimodal generative systems.

2402.11120

2024-02-16T22:48:38Z

DART: A Principled Approach to Adversarially Robust Unsupervised Domain Adaptation

Distribution shifts and adversarial examples are two major challenges for deploying machine learning models. While these challenges have been studied individually, their combination is an important topic that remains relatively under-explored. In this work, we study the problem of adversarial robustness under a common setting of distribution shift - unsupervised domain adaptation (UDA). Specifically, given a labeled source domain $D_S$ and an unlabeled target domain $D_T$ with related but different distributions, the goal is to obtain an adversarially robust model for $D_T$. The absence of target domain labels poses a unique challenge, as conventional adversarial robustness defenses cannot be directly applied to $D_T$. To address this challenge, we first establish a generalization bound for the adversarial target loss, which consists of (i) terms related to the loss on the data, and (ii) a measure of worst-case domain divergence. Motivated by this bound, we develop a novel unified defense framework called Divergence Aware adveRsarial Training (DART), which can be used in conjunction with a variety of standard UDA methods; e.g., DANN [Ganin and Lempitsky, 2015]. DART is applicable to general threat models, including the popular $\ell_p$-norm model, and does not require heuristic regularizers or architectural changes. We also release DomainRobust: a testbed for evaluating robustness of UDA models to adversarial attacks. DomainRobust consists of 4 multi-domain benchmark datasets (with 46 source-target pairs) and 7 meta-algorithms with a total of 11 variants. Our large-scale experiments demonstrate that on average, DART significantly enhances model robustness on all benchmarks compared to the state of the art, while maintaining competitive standard accuracy. The relative improvement in robustness from DART reaches up to 29.2% on the source-target domain pairs considered.