# ReMATF: Recurrent Motion-Adaptive Multi-scale Turbulence Mitigation for Dynamic Scenes

arXiv: https://arxiv.org/abs/2605.21440 · HTML: https://mlsanity.com/paper/2605.21440/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 22 of its 37 references that are in the map.
- First posted on arXiv 20 May 2026.

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

- [Deformable Convolutional Networks](https://mlsanity.com/paper/1703.06211/) (2017)
- [Exploring CLIP for Assessing the Look and Feel of Images](https://mlsanity.com/paper/2207.12396/) (2022)
- [VRT: A Video Restoration Transformer](https://mlsanity.com/paper/2201.12288/) (2022)
- [Recurrent Video Restoration Transformer with Guided Deformable Attention](https://mlsanity.com/paper/2206.02146/) (2022)
- [AT-DDPM: Restoring Faces degraded by Atmospheric Turbulence using Denoising Diffusion Probabilistic Models](https://mlsanity.com/paper/2208.11284/) (2022)
- [Imaging through the Atmosphere using Turbulence Mitigation Transformer](https://mlsanity.com/paper/2207.06465/) (2022)
- [Real-world Video Deblurring: A Benchmark Dataset and An Efficient Recurrent Neural Network](https://mlsanity.com/paper/2106.16028/) (2021)
- [Physics-Driven Turbulence Image Restoration with Stochastic Refinement](https://mlsanity.com/paper/2307.10603/) (2023)
- [Spatio-Temporal Turbulence Mitigation: A Translational Perspective](https://mlsanity.com/paper/2401.04244/) (2024)
- [LTT-GAN: Looking Through Turbulence by Inverting GANs](https://mlsanity.com/paper/2112.02379/) (2021)
- [Atmospheric Turbulence Removal with Complex-Valued Convolutional Neural Network](https://mlsanity.com/paper/2204.06989/) (2022)
- [Atmospheric turbulence mitigation for sequences with moving objects using recursive image fusion](https://mlsanity.com/paper/1808.03550/) (2018)
- [Atmospheric turbulence removal using convolutional neural network](https://mlsanity.com/paper/1912.11350/) (2019)
- [TurbuGAN: An Adversarial Learning Approach to Spatially-Varying Multiframe Blind Deconvolution with Applications to Imaging Through Turbulence](https://mlsanity.com/paper/2203.06764/) (2022)
- [Deep Learning Techniques for Atmospheric Turbulence Removal: A Review](https://mlsanity.com/paper/2409.14587/) (2024)
- [Learning Phase Distortion with Selective State Space Models for Video Turbulence Mitigation](https://mlsanity.com/paper/2504.02697/) (2025)
- [Long-range Turbulence Mitigation: A Large-scale Dataset and A Coarse-to-fine Framework](https://mlsanity.com/paper/2407.08377/) (2024)
- [Turb-Seg-Res: A Segment-then-Restore Pipeline for Dynamic Videos with Atmospheric Turbulence](https://mlsanity.com/paper/2404.13605/) (2024)
- [DeTurb: Atmospheric Turbulence Mitigation with Deformable 3D Convolutions and 3D Swin Transformers](https://mlsanity.com/paper/2407.20855/) (2024)
- [MAMAT: 3D Mamba-Based Atmospheric Turbulence Removal and its Object Detection Capability](https://mlsanity.com/paper/2503.17700/) (2025)
- [DMAT: An End-to-End Framework for Joint Atmospheric Turbulence Mitigation and Object Detection](https://mlsanity.com/paper/2507.04323/) (2025)
- [RMFAT: Recurrent Multi-scale Feature Atmospheric Turbulence Mitigator](https://mlsanity.com/paper/2508.11409/) (2025)

## Abstract (arXiv)

Atmospheric turbulence severely degrades video quality by introducing distortions such as geometric warping, blur, and temporal flickering, posing significant challenges to both visual clarity and temporal consistency. Current state-of-the-art methods are based on transformer, 3D architectures and require multi-frame input, but their large computational cost and memory usage limit real-time deployment, especially in resource-constrained scenarios. In this work, we propose ReMATF, a lightweight recurrent framework that restores videos using only two frames at a time while preserving spatial detail and temporal stability. ReMATF combines a multi-scale encoder-decoder with temporal warping and a motion-adaptive temporal fusion module that performs per-pixel fusion between the warped previous output and the current prediction to enhance coherence without enlarging the temporal window. This design reduces flicker, sharpens details, and remains efficient. Experiments on synthetic and real turbulence datasets show consistent improvements in PSNR/SSIM and perceptual quality (LPIPS), along with substantially faster inference than multi-frame transformer baselines, making ReMATF suitable turbulence mitigation in resource-constrained scenarios.