发表机构
University of Southern California（南加州大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文提出基于随机Hessian估计的IDQD-RDO方法，将全参考图像质量度量近似为逐块可计算的二次失真，在VVC中实现14.2-36.7%的BD率节省，无需解码器改动。
AI 中文摘要
基于块的视频编解码器通过优化率失真权衡，根据输入选择编码参数。传统的失真选择——平方误差和（SSE）简化了参数选择：SSE是逐块SSE之和，因此率失真优化（RDO）可以独立处理各块。相比之下，诸如MS-SSIM或LPIPS等全参考图像质量评估（FR-IQA）度量通常比SSE更符合人类视觉系统，但它们不能用于环路内：它们不能逐块分解，并且通常需要完全解码的图像作为输入。基于现有的度量二次化结果，我们将一类广泛的FR-IQA度量近似为输入相关的二次失真（IDQD），其二次型矩阵由在源视频处评估的度量的Hessian矩阵导出。为了使失真可逐块计算，我们提出了Hessian矩阵的两种近似：1）保留块对角部分，2）仅保留其对角部分。我们为两者提出了估计器，这些估计器仅需与通过自动微分获得的Hessian矩阵进行矩阵-向量乘积。在VVC中针对Kodak和CLIC的五个度量上，IDQD-RDO在目标度量下实现了14.2-36.7%的BD率节省，无需更改解码器，并产生10-30%的编码复杂度开销。
英文摘要
Block-based video codecs select coding parameters based on the input by optimizing a rate-distortion trade-off. The conventional distortion choice, the sum of squared errors (SSE), simplifies parameter selection: the SSE is the sum of block-wise SSEs, so rate-distortion optimization (RDO) can treat blocks independently. Alternatively, full-reference image quality assessment (FR-IQA) metrics such as MS-SSIM or LPIPS often align better with the human visual system than SSE, but they cannot be used in-loop: they do not decompose block-wise and typically require the fully decoded image as input. Building on existing results in metric quadratization, we approximate a broad class of FR-IQA metrics by an input-dependent quadratic distortion (IDQD), whose quadratic form matrix is derived from the Hessian of the metric evaluated at the source video. To make the distortion computable block-wise, we propose two approximations of the Hessian matrix: 1) keeping the block-diagonal, and 2) keeping only its diagonal. We propose estimators for both that require only matrix-vector products with the Hessian obtained by automatic differentiation. Across five metrics for Kodak and CLIC in VVC, IDQD-RDO achieves 14.2-36.7 % BD-rate savings under the target metric with no decoder changes and incurs 10-30 % encoding complexity overhead.
