用神经网络预处理视频，让H.264编码更清晰，兼容所有播放器。
Kelvin v1.0: A Neural Pre-Encoder for H.264: A standards-compliant learned preprocessor with -27.62% BD-VMAF on UVG
- 在H.264前加轻量神经预编码器，自适应调整像素，提升感知质量。
- 在UVG基准上平均提升27.62%的视觉质量（BD-VMAF），7个序列全胜。
- 专为必须用H.264的场景设计，不改变原有解码兼容性。
Kelvin v1.0 是一个轻量级的神经预编码器，置于未修改的 libx264 编码器之前。它对像素进行内容自适应调整，每通道变动范围限定在 ±1/255 内，使编码器将比特分配到最影响感知质量的位置，同时输出符合标准的 H.264 比特流，兼容所有现有解码器、播放器和 CDN。在7个1080p UVG序列基准上，相对于 baseline libx264（preset medium），Kelvin v1.0 实现了 -27.62% 的平均 BD-VMAF（7/7 获胜）和 -5.18% 的 BD-VMAF-NEG（6/7 获胜）。在包含30个序列的 MCL-JCV 公共数据集（28个训练未见）上，同一模型在28个片段中取得更好的 BD-VMAF；移除两个可诊断失败后，平均为 -27.70% BD-VMAF 与 -5.37% BD-VMAF-NEG，与 UVG 结果一致（误差≤1个百分点）。核心工程挑战是 H.264 的不可微性：本文提出一种混合编解码代理，结合校准的可微分率估计算子（与真实 libx264 的 bits-per-pixel 相关性 Spearman rho = 0.986）和基于真实编码输出训练的 U-Net 失真代理。我们公开了完整逐序列率失真数据、MCL-JCV 上的命名故障模式分类（速率下限违规、分布偏移、指标饱和）、五种基线对照组（hqdn3d, unsharp, -tune psnr, -tune ssim, x265 medium），并坦诚声明：x265 medium 在相同数据集上各项指标均优于 Kelvin。因此，Kelvin 专为必须使用 H.264 的工作负载而设计，而非主动选择。
原文摘要 · Abstract (English)
Kelvin is a lightweight learned pre-encoder that sits in front of an unmodified libx264 encoder. It applies content-adaptive pixel adjustments, bounded at +/-1/255 per channel, so that the encoder allocates bits where they matter most perceptually, while emitting a standard H.264 bitstream compatible with every existing decoder, player, and CDN. On the seven-sequence 1080p UVG benchmark, Kelvin v1.0 achieves a mean BD-VMAF of -27.62% (7 of 7 wins) and BD-VMAF-NEG of -5.18% (6 of 7 wins) relative to baseline libx264 at preset medium. On the 30-sequence MCL-JCV public set (28 unseen by training), the same checkpoint wins on 28 of 30 clips by BD-VMAF; with the two diagnosable failures removed the mean is -27.70% BD-VMAF and -5.37% BD-VMAF-NEG, consistent with UVG to within one percentage point. A central engineering challenge is the non-differentiability of H.264: we describe a hybrid codec proxy that combines a calibrated differentiable rate estimator (Spearman rho = 0.986 vs. real libx264 bits-per-pixel) with a U-Net distortion proxy trained on real encoder outputs. We publish full per-sequence rate-distortion data, a named failure-mode taxonomy on MCL-JCV (rate-floor violation, distribution shift, metric saturation), a five-baseline sanity panel (hqdn3d, unsharp, -tune psnr, -tune ssim, x265 medium), and honest positioning: x265 medium beats Kelvin on every metric on the same corpus. Kelvin is therefore designed for workloads where remaining on H.264 is a constraint rather than a choice.
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。