metadata

title: Super Resolution
emoji: 🏢
colorFrom: indigo
colorTo: red
sdk: streamlit
sdk_version: 1.29.0
app_file: app.py
pinned: false
license: apache-2.0

Super Resolution Deep Learning Project - Group 11

Members

Nguyen Ba Thiem (20214931)
Doan The Vinh (20210940)
Pham Quang Tung (20210919)
Nguyen Hai Long (20214911)
Billy Tom Herrmann (2023T033)

Introduction

This project aims to develop a deep learning model for enhancing the resolution of images, known as Super Resolution. Our approach utilizes state-of-the-art techniques in deep learning to upscale images while preserving quality.

Installation

To clone this report, make sure you follow steps below to download large files.

git lfs install
git clone https://huggingface.co/spaces/thiemcun203/super-resolution

Before running the application, ensure you have the necessary packages installed:

pip install -r requirements.txt

Running the Application

To run the web application, use the following command:

streamlit run app.py

We also deployed this app on hugging face, you can check it by the following link. However, because of the limit of resources, the time to inference will be longer than running on local machine Super Resolution GUI online

Inference and Results

To run inference and check results, follow the detailed steps provided in our Kaggle notebook:

Infer Super Resolution

Folders Structure

`models/`

Contains all models with checkpoints, file architecture, and training files necessary for the super-resolution process.

`images/`

Use this folder to test the application with some images. Place your low-resolution images here and run the application to see the enhanced results.

Acknowledgements

We would like to express our sincerest gratitude to our lecturer, Prof. Nguyen Hung Son, for his lectures and guidance throughout this semester. We would also like to extend our thanks to Mr. Nguyen The Minh Duc and Mr. Ho Minh Khoi, the Teaching Assistants of the class, for their support throughout the course, as well as their invaluable advice in approaching the project.

Group 11 - Deep Learning for Image Super-Resolution