Official Pytorch implementation of Diffusion-based Bit-depth Expansion.
Riyu Lu, Lingyu Zhu, Baoliang Chen, Xiaopeng Fan, Shiqi Wang
[PDF] [Poster] [Presentation]
Diffusion-based generative models have achieved remarkable success across a variety of applications. However, the potential application for bit-depth expansion has not been extensively studied. This paper introduces a wavelet-based diffusion model for the bit-depth expansion task. In this method, the image is first decomposed into low and high-frequency components via wavelet transformation. This decomposition allows for targeted processing by specialized modules and reduces computational complexity by lowering the image resolution. The low-frequency component is processed in both the forward diffusion and reverse denoising stages. Meanwhile, the high-frequency components are filtered by the High Frequency Denoising Filter (HFDF) to eliminate noise and artifacts. Finally, the low and high-frequency components are recombined into a predicted high-bit-depth image through inverse wavelet transformation. Experimental results demonstrate the superiority of the proposed method in producing perceptually compelling outputs that outperform previous methods.
This repository is still under active construction:
- Release training and testing codes
- Release pretrained models
- Clean the code
- Lingyu Zhu: [email protected]
- Riyu Lu: [email protected]
If you find our work helpful, please consider citing:
@INPROCEEDINGS{10743597,
author={Lu, Riyu and Zhu, Lingyu and Chen, Baoliang and Fan, Xiaopeng and Wang, Shiqi},
booktitle={2024 IEEE 26th International Workshop on Multimedia Signal Processing (MMSP)},
title={Diffusion-Based Bit-Depth Expansion},
year={2024}}