DA-Flow: Degradation-Aware Optical Flow Estimation with Diffusion Models

TL;DR

DA-Flow introduces a degradation-aware optical flow estimation method that leverages diffusion model features with spatio-temporal attention, significantly improving accuracy on corrupted videos.

Contribution

The paper proposes a novel hybrid architecture combining diffusion model features with convolutional features for degradation-aware optical flow estimation.

Findings

DA-Flow outperforms existing methods on corrupted video benchmarks.

Diffusion model features exhibit zero-shot correspondence capabilities.

Spatio-temporal attention enhances feature robustness to corruptions.

Abstract

Optical flow models trained on high-quality data often degrade severely when confronted with real-world corruptions such as blur, noise, and compression artifacts. To overcome this limitation, we formulate Degradation-Aware Optical Flow, a new task targeting accurate dense correspondence estimation from real-world corrupted videos. Our key insight is that the intermediate representations of image restoration diffusion models are inherently corruption-aware but lack temporal awareness. To address this limitation, we lift the model to attend across adjacent frames via full spatio-temporal attention, and empirically demonstrate that the resulting features exhibit zero-shot correspondence capabilities. Based on this finding, we present DA-Flow, a hybrid architecture that fuses these diffusion features with convolutional features within an iterative refinement framework. DA-Flow substantially outperforms existing optical flow methods under severe degradation across multiple benchmarks.