提出可微向量量化框架，实现图像压缩的端到端率失真优化。
Differentiable Vector Quantization for Rate-Distortion Optimization of Generative Image Compression
- 通过可微松弛代码本分布，让熵损失直接调节潜在先验。
- 在极低码率下性能优越，参数量少，比特率降低超37%。
- 适合研究高效图像压缩与生成式编码的学者使用。
在存储和带宽受限的背景下，极低码率图像压缩愈发重要。尽管向量量化（VQ）能保持良好结构保真度，但现有方法因表征学习与熵建模脱节，缺乏联合率失真（RD）优化的理论机制。本文提出RDVQ，一种统一框架，通过代码本分布的可微松弛，实现基于VQ压缩的端到端RD优化，使熵损失能直接塑造潜在先验。进一步设计了自回归熵模型，支持精确熵建模与测试时码率控制。大量实验表明，RDVQ以轻量架构在极低码率下表现优异，感知质量竞争力强且参数显著减少。相比RDEIC，RDVQ在DIV2K-val上于DISTS指标降低最高75.71%比特率，在LPIPS指标降低37.63%。除实证提升外，RDVQ引入了熵约束下的VQ形式化，为图像标记化与压缩的统一视角提供可能。代码将公开于https://github.com/CVL-UESTC/RDVQ。
原文摘要 · Abstract (English)
The rapid growth of visual data under stringent storage and bandwidth constraints makes extremely low-bitrate image compression increasingly important. While Vector Quantization (VQ) offers strong structural fidelity, existing methods lack a principled mechanism for joint rate-distortion (RD) optimization due to the disconnect between representation learning and entropy modeling. We propose RDVQ, a unified framework that enables end-to-end RD optimization for VQ-based compression via a differentiable relaxation of the codebook distribution, allowing the entropy loss to directly shape the latent prior. We further develop an autoregressive entropy model that supports accurate entropy modeling and test-time rate control. Extensive experiments demonstrate that RDVQ achieves strong performance at extremely low bitrates with a lightweight architecture, attaining competitive or superior perceptual quality with significantly fewer parameters. Compared with RDEIC, RDVQ reduces bitrate by up to 75.71% on DISTS and 37.63% on LPIPS on DIV2K-val. Beyond empirical gains, RDVQ introduces an entropy-constrained formulation of VQ, highlighting the potential for a more unified view of image tokenization and compression. The code will be available at https://github.com/CVL-UESTC/RDVQ.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。