DUO-VSR：面向单步视频超分辨率的双流蒸馏
DUO-VSR: Dual-Stream Distillation for One-Step Video Super-Resolution
把分布匹配蒸馏与对抗监督（RFS-GAN）结合成双流蒸馏，可将任意多步扩散 VSR 模型压缩到单步推理；其渐进式引导蒸馏初始化与偏好引导精修流程，也可迁移到其他扩散复原任务（去噪、去模糊）的少步加速上。
原文依据 we propose DUO-VSR, a three-stage framework centered on a DUal-Stream Distillation strategy
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论基于扩散模型的视频超分辨率（VSR）能够获得出色的保真度，但采样开销高得难以承受。尽管分布匹配蒸馏（DMD）可将扩散模型加速至单步生成，但直接将其应用于 VSR 会导致训练不稳定以及监督不足、质量下降的问题。为解决这些问题，我们提出 DUO-VSR，这是一个以双流蒸馏策略为核心的三阶段框架，将分布匹配与对抗监督相结合，用于单步 VSR。我们首先采用渐进式引导蒸馏初始化，通过保持轨迹的蒸馏来稳定后续训练。随后，我们引入双流蒸馏策略，联合优化 DMD 流与真实-伪造评分特征生成对抗网络（RFS-GAN）流，后者利用来自真实评分模型与伪造评分模型的特征提供互补的对抗监督。最后，偏好引导精修使学生模型与感知质量偏好对齐。大量实验表明，DUO-VSR 在视觉质量与效率上均优于以往的单步 VSR 方法。
英文原文摘要 1129 字符
Diffusion-based video super-resolution (VSR) achieves remarkable fidelity but suffers from prohibitive sampling cost. While distribution matching distillation (DMD) accelerates diffusion models to one-step generation, directly applying it to VSR leads to training instability and degraded, insufficient supervision.To address these issues, we propose DUO-VSR, a three-stage framework centered on a DUal-Stream Distillation strategy that integrates distribution matching and adversarial supervision for One-step VSR.We first adopt a Progressive Guided Distillation Initialization to stabilize subsequent training through trajectory-preserving distillation.We then introduce a Dual-Stream Distillation Strategy to jointly optimize DMD and Real-Fake Score Feature GAN (RFS-GAN) streams, with the latter providing complementary adversarial supervision using features from both real and fake score models.Finally, a Preference-Guided Refinement aligns the student with perceptual quality preferences.Comprehensive experiments demonstrate that DUO-VSR achieves superior visual quality and efficiency over previous one-step VSR methods.