Summary of Multiple Latent Space Mapping For Compressed Dark Image Enhancement, by Yi Zeng et al.
Multiple Latent Space Mapping for Compressed Dark Image Enhancement
by Yi Zeng, Zhengning Wang, Yuxuan Liu, Tianjiao Zeng, Xuhang Liu, Xinglong Luo, Shuaicheng Liu, Shuyuan Zhu, Bing Zeng
First submitted to arxiv on: 12 Mar 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Image and Video Processing (eess.IV)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper tackles the problem of enhancing dark images that have been compressed before storage or transmission. Current methods struggle with this task, amplifying artifacts hidden in the dark regions and resulting in uncomfortable visual effects. To address this issue, the authors propose a novel latent mapping network based on variational auto-encoder (VAE) to enhance compressed dark images while avoiding compression artifacts amplification. The method involves training two multi-level VAEs to project compressed dark images and normal-light images into their respective latent spaces, and then using a latent mapping network to transform features from the compressed dark space to the normal-light space. The proposed method achieves state-of-the-art performance in compressed dark image enhancement. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper is about improving how we make dark pictures look better when they were shrunk before being stored or sent online. Right now, our best methods don’t do a good job with these kinds of pictures because they make the tiny details that are already lost get even worse. The authors have come up with a new way to fix this problem using a special kind of computer program called a variational auto-encoder (VAE). Their method uses two different VAEs to look at both shrunk and normal-sized pictures, and then it uses those results to make the dark picture look better while keeping all the good parts. The new method is really good at making dark pictures look better. |
Keywords
» Artificial intelligence » Encoder
