Dataset Description

We introduce a challenging dataset for identifying machine parts from real photos, featuring images of 102 parts from a labeling machine. This dataset was developed with the complexity of real-world scenarios in mind and highlights the complexity of distinguishing between closely related classes, providing an opportunity to improve domain adaption methods. The dataset includes 3,264 CAD-rendered images (32 per part) and 6,146 real images (6 to 137 per part) for UDA and testing. Rendered images were produced using a Blender-based pipeline with environment maps, lights, and virtual cameras arranged to ensure varied mesh orientations. We also use material metadata and apply one of 21 texture materials to the objects. We render all images at 512x512 pixels. The real photo set consists of raw images captured under varying conditions using different cameras, including varied lighting, backgrounds, and environmental factors.

Update:

Fix material issues for some objects. (real was black steel but synth was natural steel)
Add train & test estimated depth data from ZoeDepth
Add unprocessed (uncropped) test image data with bounding box labels
Add depth data exported from render pipeline (blender) via compositing graph. (raw EXR & normalized PNG)
Add training images including ControlNet generated wood backgrounds
Add training images including ControlNet generted hands
Add training images processed by T2i-Adapter Style Transfer

Download

Download zipped dataset

Licensing Information

CC BY-NC 4.0 Deed

Citation Information

Please cite our work if you use the data set.

@InProceedings{10.1007/978-3-031-74640-6_33,
author="Ritter, Dennis
and Hemberger, Mike
and H{\"o}nig, Marc
and Stopp, Volker
and Rodner, Erik
and Hildebrand, Kristian",
editor="Meo, Rosa
and Silvestri, Fabrizio",
title="CAD Models to Real-World Images: A Practical Approach to Unsupervised Domain Adaptation in Industrial Object Classification",
booktitle="Machine Learning and Principles and Practice of Knowledge Discovery in Databases",
year="2025",
publisher="Springer Nature Switzerland",
address="Cham",
pages="399--415",
abstract="In this paper, we systematically analyze unsupervised domain adaptation pipelines for object classification in a challenging industrial setting. In contrast to standard natural object benchmarks existing in the field, our results highlight the most important design choices when only category-labeled CAD models are available but classification needs to be done with real-world images. Our domain adaptation pipeline achieves SoTA performance on the VisDA benchmark, but more importantly, drastically improves recognition performance on our new open industrial dataset comprised of 102 mechanical parts. We conclude with a set of guidelines that are relevant for practitioners needing to apply state-of-the-art unsupervised domain adaptation in practice. Our code is available at https://github.com/dritter-bht/synthnet-transfer-learning.",
isbn="978-3-031-74640-6"
}