Summary of Misc: Ultra-low Bitrate Image Semantic Compression Driven by Large Multimodal Model, By Chunyi Li et al.
MISC: Ultra-low Bitrate Image Semantic Compression Driven by Large Multimodal Model
by Chunyi Li, Guo Lu, Donghui Feng, Haoning Wu, Zicheng Zhang, Xiaohong Liu, Guangtao Zhai, Weisi Lin, Wenjun Zhang
First submitted to arxiv on: 26 Feb 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 proposes a novel image compression method called Multimodal Image Semantic Compression (MISC) that balances consistency with the ground truth and perceptual quality at ultra-low bitrate. By leveraging the Large Multimodal Model (LMM), MISC consists of an LMM encoder, map encoder, image encoder, and decoder to compress images and reconstruct them accurately. Experimental results demonstrate that MISC can effectively compress both traditional Natural Sense Images (NSIs) and AI-Generated Images (AIGIs), achieving optimal consistency and perception while reducing bitrate by 50%. The method has strong potential applications in next-generation storage and communication. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper introduces a new way to shrink images without losing important details. It uses a special kind of model called the Large Multimodal Model (LMM) to find the most important parts of an image, and then compresses those parts into a very small file. This method works for both normal pictures and AI-generated ones. By shrinking files in this way, it could make storing and sharing images faster and easier. |
Keywords
» Artificial intelligence » Decoder » Encoder
