提出低复杂度神经压缩器，兼顾率失真感知三者平衡，实现理论最优性能。
Optimal Neural Compressors for the Rate-Distortion-Perception Tradeoff
- 采用格子编码提升空间打包效率，共享抖动引入随机性增强重构质量。
- 在无限与零共享随机性两种场景下，均达到理论最优的率-失真-感知权衡。
- 实验验证：更多共享随机性和更好格子打包能显著提升压缩性能。
近年来，神经压缩研究聚焦于率-失真-感知（RDP）权衡，其中感知约束要求源分布与重建分布之间的统计差异较小。现有理论描述了RDP最优压缩器的特性，但未提供可构造且低复杂度的解决方案。经典率失真理论表明最优压缩器应高效填充空间，而RDP理论进一步指出，为实现最优性，编码器与解码器之间可能需要无限共享随机性。本文提出一种低复杂度的神经压缩器，通过格子编码实现高空间打包效率，并利用格子单元上的共享抖动引入共享随机性。针对两种关键情形——无限共享随机性与零共享随机性——我们分析了所提压缩器实现的RDP权衡，证明其在两种情况下均达到最优。实验上，我们在合成数据和真实世界数据上研究了格子编码与随机性这两个设计成分对性能的影响，发现性能随共享随机性增加和格子打包效果提升而改善。
原文摘要 · Abstract (English)
Recent efforts in neural compression have focused on the rate-distortion-perception (RDP) tradeoff, where the perception constraint ensures the source and reconstruction distributions are close in terms of a statistical divergence. Theoretical work on RDP describes properties of RDP-optimal compressors without providing constructive and low complexity solutions. While classical rate distortion theory shows that optimal compressors should efficiently pack space, RDP theory additionally shows that infinite randomness shared between the encoder and decoder may be necessary for RDP optimality. In this paper, we propose neural compressors that are low complexity and benefit from high packing efficiency through lattice coding and shared randomness through shared dithering over the lattice cells. For two important settings, namely infinite shared and zero shared randomness, we analyze the RDP tradeoff achieved by our proposed neural compressors and show optimality in both cases. Experimentally, we investigate the roles that these two components of our design, lattice coding and randomness, play in the performance of neural compressors on synthetic and real-world data. We observe that performance improves with more shared randomness and better lattice packing.
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。