Summary of Exploring Compressed Image Representation As a Perceptual Proxy: a Study, by Chen-hsiu Huang and Ja-ling Wu
Exploring Compressed Image Representation as a Perceptual Proxy: A Study
by Chen-Hsiu Huang, Ja-Ling Wu
First submitted to arxiv on: 14 Jan 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Machine Learning (cs.LG); 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 proposed end-to-end learned image compression codec combines an analysis transform with object classification task training, allowing for joint optimization. This approach achieves comparable accuracy to custom-designed DNN-based quality metrics in predicting human perceptual distance judgments. Additionally, the study explores various neural encoders and demonstrates the effectiveness of using the analysis transform as a perceptual loss network for tasks beyond quality assessments. The off-the-shelf encoder proves proficient in perceptual modeling without requiring an additional VGG network. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This research proposes a new way to compress images by training two things together: an analysis transform that helps us understand what’s in the picture, and a task to classify objects. This makes it better at predicting how good the compressed image looks compared to other methods. The study also shows that using this analysis transform can help with other tasks like image quality assessment without needing extra networks. |
Keywords
* Artificial intelligence * Classification * Encoder * Optimization
