FluidRain: 不可压缩雨流作为环中环视频去雨的注意力偏置
FluidRain: Incompressible Rain Flow as an Attention Bias for Loop-in-Loop Video Deraining
浏览论文内容
中文总结 AI 辅助
提出FluidRain，利用无散度雨流引导环中环注意力，实现轻量级视频去雨，仅0.80M参数，在四个基准上媲美大型模型，并引入RainSyn-Gust和物理无参考指标。
中文摘要 AI 辅助
现有的视频去雨方法通常通过显式对齐或隐式时空聚合来利用相邻帧。显式对齐依赖于准确的运动估计，在密集雨条件下可能变得不可靠，而隐式聚合避免了对齐，但缺乏对雨的方向性和时间连贯性结构的显式指导。这就在可靠的时间聚合与雨的显式运动建模之间留下了空白。为解决这些局限性，我们提出了FluidRain，一种轻量级视频去雨器，它使用无散度雨流来引导跨尺度和相邻帧的环中环注意力。受流体力学启发，我们将雨运动建模为图像空间中的无散度流，并用它来组织多尺度和时间聚合。具体来说，FluidRain首先为每帧估计一个雨流场，并将其投影到无散度子空间。由此产生的流引导窗口注意力沿雨条纹方向，使得相邻帧无需显式对齐即可聚合。由于雨流结构在尺度和邻近帧间得以保留，环中环在两个维度上重用相同的注意力算子，从而得到一个仅0.80M参数的三帧模型。在四个基准上的实验表明，FluidRain与规模大得多的恢复模型相比仍具有竞争力。我们进一步研究了时间证据如何随不同输入视图扩展。为了评估模型在帧间雨运动变化时是否保持可靠，我们引入了RainSyn-Gust，它将雨条纹方向的受控变化注入现有基准。我们还开发了一种基于物理的无参考指标，无需干净目标即可评估真实雨去除效果。
英文摘要
Existing video deraining methods typically exploit neighboring frames through either explicit alignment or implicit spatiotemporal aggregation. Explicit alignment relies on accurate motion estimation, which can become unreliable under dense rain, while implicit aggregation avoids alignment but lacks explicit guidance on the directional and temporally coherent structure of rain. This leaves a gap between reliable temporal aggregation and explicit modeling of rain motion. To address these limitations, we propose FluidRain, a lightweight video derainer that uses divergence-free rain flow to guide Loop-in-Loop attention across scales and neighboring frames. Motivated by fluid mechanics, we model rain motion as a divergence-free image-space flow and use it to organize multi-scale and temporal aggregation. Specifically, FluidRain first estimates a rain-flow field for each frame and projects it onto the divergence-free subspace. The resulting flow steers window attention along rain streaks, enabling neighboring frames to be aggregated without explicit alignment. Since rain-flow structure is preserved across scales and nearby frames, Loop-in-Loop reuses the same attention operator across both dimensions, resulting in a three-frame model with only 0.80M parameters. Experiments on four benchmarks show that FluidRain remains competitive with substantially larger restoration models. We further examine how temporal evidence scales with different input views. To evaluate whether the model remains reliable when rain motion changes across frames, we introduce RainSyn-Gust, which injects controlled changes in rain-streak direction into existing benchmarks. We also develop a physics-based no-reference metric that evaluates real-rain removal without requiring clean targets.
发表机构
- Shandong University（山东大学）
- Central South University（中南大学）
- Nanjing University of Science and Technology（南京理工大学）
- Beijing University of Posts and Telecommunications（北京邮电大学）
- National University of Defense Technology（国防科技大学）
机构由 AI 辅助整理，请以论文原文为准。
