MISC: Ultra-low Bitrate Image Semantic Compression Driven by Large Multimodal Model

TL;DR

MISC leverages large multimodal models to achieve ultra-low bitrate image compression, balancing fidelity and perceptual quality, and is effective for both natural and AI-generated images, saving 50% bitrate.

Contribution

Introduces MISC, a novel semantic compression method using LMMs, enabling ultra-low bitrate image compression with maintained quality and broad applicability.

Findings

Achieves 50% bitrate reduction while maintaining quality.

Effective for both natural and AI-generated images.

Balances ground truth fidelity and perceptual quality.

Abstract

With the evolution of storage and communication protocols, ultra-low bitrate image compression has become a highly demanding topic. However, existing compression algorithms must sacrifice either consistency with the ground truth or perceptual quality at ultra-low bitrate. In recent years, the rapid development of the Large Multimodal Model (LMM) has made it possible to balance these two goals. To solve this problem, this paper proposes a method called Multimodal Image Semantic Compression (MISC), which consists of an LMM encoder for extracting the semantic information of the image, a map encoder to locate the region corresponding to the semantic, an image encoder generates an extremely compressed bitstream, and a decoder reconstructs the image based on the above information. Experimental results show that our proposed MISC is suitable for compressing both traditional Natural Sense Images (NSIs) and emerging AI-Generated Images (AIGIs) content. It can achieve optimal consistency and perception results while saving 50% bitrate, which has strong potential applications in the next generation of storage and communication. The code will be released on https://github.com/lcysyzxdxc/MISC.