# FluidRain: Incompressible Rain Flow as an Attention Bias for Loop-in-Loop Video Deraining

arXiv: https://arxiv.org/abs/2609.29006 · HTML: https://mlsanity.com/paper/2609.29006/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 35 references that are in the map.
- First posted on arXiv 24 Sep 2026.

## Builds on (most cited first, 22 of 22)

- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [SwinIR: Image Restoration Using Swin Transformer](https://mlsanity.com/paper/2108.10257/) (2021)
- [Restormer: Efficient Transformer for High-Resolution Image Restoration](https://mlsanity.com/paper/2111.09881/) (2021)
- [Simple Baselines for Image Restoration](https://mlsanity.com/paper/2204.04676/) (2022)
- [Universal Transformers](https://mlsanity.com/paper/1807.03819/) (2018)
- [Deep Joint Rain Detection and Removal from a Single Image](https://mlsanity.com/paper/1609.07769/) (2016)
- [BasicVSR++: Improving Video Super-Resolution with Enhanced Propagation and Alignment](https://mlsanity.com/paper/2104.13371/) (2021)
- [Deeply-Recursive Convolutional Network for Image Super-Resolution](https://mlsanity.com/paper/1511.04491/) (2015)
- [Learning A Sparse Transformer Network for Effective Image Deraining](https://mlsanity.com/paper/2303.11950/) (2023)
- [Relaxed Recursive Transformers: Effective Parameter Sharing with Layer-wise LoRA](https://mlsanity.com/paper/2410.20672/) (2024)
- [RainMamba: Enhanced Locality Learning with State Space Models for Video Deraining](https://mlsanity.com/paper/2407.21773/) (2024)
- [Bidirectional Multi-Scale Implicit Neural Representations for Image Deraining](https://mlsanity.com/paper/2404.01547/) (2024)
- [Robust Video Content Alignment and Compensation for Rain Removal in a CNN Framework](https://mlsanity.com/paper/1803.10433/) (2018)
- [From Sky to the Ground: A Large-scale Benchmark and Simple Baseline Towards Real Rain Removal](https://mlsanity.com/paper/2308.03867/) (2023)
- [Semi-Supervised Video Deraining with Dynamical Rain Generator](https://mlsanity.com/paper/2103.07939/) (2021)
- [Enhanced Spatio-Temporal Interaction Learning for Video Deraining: A Faster and Better Framework](https://mlsanity.com/paper/2103.12318/) (2021)
- [How Much Is One Recurrence Worth? Iso-Depth Scaling Laws for Looped Language Models](https://mlsanity.com/paper/2604.21106/) (2026)
- [Video Adverse-Weather-Component Suppression Network via Weather Messenger and Adversarial Backpropagation](https://mlsanity.com/paper/2309.13700/) (2023)
- [Controllable Weather Synthesis and Removal with Video Diffusion Models](https://mlsanity.com/paper/2505.00704/) (2025)
- [Semi-Supervised State-Space Model with Dynamic Stacking Filter for Real-World Video Deraining](https://mlsanity.com/paper/2505.16811/) (2025)
- [DeLiVR: Differential Spatiotemporal Lie Bias for Efficient Video Deraining](https://mlsanity.com/paper/2509.21719/) (2025)
- [ASF-Net: Robust Video Deraining via Temporal Alignment and Online Adaptive Learning](https://mlsanity.com/paper/2309.00956/) (2023)

## Abstract (arXiv)

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.