Summary of Learned Image Compression with Text Quality Enhancement, by Chih-yu Lai et al.
Learned Image Compression with Text Quality Enhancement
by Chih-Yu Lai, Dung Tran, Kazuhito Koishida
First submitted to arxiv on: 13 Feb 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Machine Learning (cs.LG)
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 This research proposes a novel approach to minimize text distortion in learned image compression at ultra-low bit-rates, particularly for screen-content images (SCI). The method involves designing a novel text logit loss function that quantifies the disparity between original and reconstructed text. Through extensive experimentation on diverse datasets using state-of-the-art algorithms, the proposed loss function demonstrates significant improvements in text quality when integrated with appropriate weighting. Notably, this approach achieves a Bjontegaard delta (BD) rate of -32.64% for Character Error Rate (CER) and -28.03% for Word Error Rate (WER) on average for two screenshot datasets. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps make images smaller while keeping the text clear. When you compress an image too much, the text might get blurry or hard to read. To fix this, researchers came up with a new way to measure how different the original and compressed text is. They tested this method on many different types of images and found that it works really well. This could be useful for people who want to send images quickly over the internet without sacrificing text quality. |
Keywords
* Artificial intelligence * Cer * Loss function
