Neural Image Compression with Text-guided Encoding for both Pixel-level and Perceptual Fidelity

TL;DR

This paper introduces a novel text-guided image compression method that maintains both high perceptual quality and pixel-level fidelity by leveraging text-adaptive encoding and joint image-text training, outperforming existing approaches.

Contribution

The authors propose a new compression framework that effectively utilizes semantic text information without relying on text-guided generative models, improving both perceptual and pixel fidelity.

Findings

Achieves superior LPIPS scores compared to baselines.

Maintains high pixel-level fidelity alongside perceptual quality.

Effective with both human- and machine-generated captions.

Abstract

Recent advances in text-guided image compression have shown great potential to enhance the perceptual quality of reconstructed images. These methods, however, tend to have significantly degraded pixel-wise fidelity, limiting their practicality. To fill this gap, we develop a new text-guided image compression algorithm that achieves both high perceptual and pixel-wise fidelity. In particular, we propose a compression framework that leverages text information mainly by text-adaptive encoding and training with joint image-text loss. By doing so, we avoid decoding based on text-guided generative models -- known for high generative diversity -- and effectively utilize the semantic information of text at a global level. Experimental results on various datasets show that our method can achieve high pixel-level and perceptual quality, with either human- or machine-generated captions. In particular, our method outperforms all baselines in terms of LPIPS, with some room for even more improvements when we use more carefully generated captions.