U$^{2}$Flow: Uncertainty-Aware Unsupervised Optical Flow Estimation

TL;DR

U$^{2}$Flow is a novel recurrent unsupervised optical flow framework that jointly estimates flow and pixel-wise uncertainty, improving robustness and interpretability without requiring ground truth data.

Contribution

It introduces a decoupled learning strategy for uncertainty supervision, an uncertainty-guided flow refinement, and a bidirectional flow fusion mechanism, advancing unsupervised optical flow estimation.

Findings

Achieves state-of-the-art results on KITTI and Sintel among unsupervised methods.

Produces highly reliable uncertainty maps alongside optical flow.

Enhances robustness in challenging regions through uncertainty-guided fusion.

Abstract

Unsupervised optical flow methods typically lack reliable uncertainty estimation, limiting their robustness and interpretability. We propose U$^{2}$Flow, the first recurrent unsupervised framework that jointly estimates optical flow and per-pixel uncertainty. The core innovation is a decoupled learning strategy that derives uncertainty supervision from augmentation consistency via a Laplace-based maximum likelihood objective, enabling stable training without ground truth. The predicted uncertainty is further integrated into the network to guide adaptive flow refinement and dynamically modulate the regional smoothness loss. Furthermore, we introduce an uncertainty-guided bidirectional flow fusion mechanism that enhances robustness in challenging regions. Extensive experiments on KITTI and Sintel demonstrate that U$^{2}$Flow achieves state-of-the-art performance among unsupervised methods while producing highly reliable uncertainty maps, validating the effectiveness of our joint estimation paradigm. The code is available at https://github.com/sunzunyi/U2FLOW.