Dynamic Scene Video Deblurring using Non-Local Attention

TL;DR

This paper introduces a novel non-local attention mechanism for video deblurring that efficiently captures spatio-temporal information without alignment, outperforming existing methods in accuracy and efficiency.

Contribution

It proposes a factorized spatio-temporal attention model that eliminates the need for alignment, improving deblurring performance and computational efficiency.

Findings

Outperforms existing fusion techniques in accuracy

Reduces computational cost compared to alignment-based methods

Demonstrates superior results on multiple datasets

Abstract

This paper tackles the challenging problem of video deblurring. Most of the existing works depend on implicit or explicit alignment for temporal information fusion which either increase the computational cost or result in suboptimal performance due to wrong alignment. In this study, we propose a factorized spatio-temporal attention to perform non-local operations across space and time to fully utilize the available information without depending on alignment. It shows superior performance compared to existing fusion techniques while being much efficient. Extensive experiments on multiple datasets demonstrate the superiority of our method.