lucs-C / FMA-Net

Official repository of FMA-Net (arXiv 2024)

FMA-Net: Flow-Guided Dynamic Filtering and Iterative Feature Refinement with Multi-Attention for Joint Video Super-Resolution and Deblurring

Geunhyuk Youk¹  Jihyong Oh^{† 2}  Munchurl Kim^{† 1}

^†Co-corresponding authors

¹Korea Advanced Institute of Science and Technology, South Korea

²Chung-Ang University, South Korea

---

This repository is the official PyTorch implementation of "FMA-Net: Flow-Guided Dynamic Filtering and Iterative Feature Refinement with Multi-Attention for Joint Video Super-Resolution and Deblurring". FMA-Net achieves state-of-the-art performance in joint video super-resolution and deblurring (VSRDB).

Please visit our project page for more visual results.

🎬 Network Architecture

Codes and pretrained models will be updated soon.

---

Reference

Geunhyuk Youk, Jihyong Oh† and Munchurl Kim† "FMA-Net: Flow-Guided Dynamic Filtering and Iterative Feature Refinement with Multi-Attention for Joint Video Super-Resolution and Deblurring", arXiv, 2024. (†: Co-corresponding authors)

@article{youk2024fmanet,
  author    = {Geunhyuk Youk and Jihyong Oh and Munchurl Kim},
  title     = {FMA-Net: Flow-Guided Dynamic Filtering and Iterative Feature Refinement with Multi-Attention for Joint Video Super-Resolution and Deblurring},
  journal   = {arXiv preprint arXiv:2401.03707},
  year      = {2024},
 }

---