UnSAMFlow: Unsupervised Optical Flow Guided by Segment Anything Model

TL;DR

UnSAMFlow is an unsupervised optical flow method that incorporates object-level information from the Segment Anything Model, improving boundary accuracy and robustness across domains.

Contribution

It introduces a novel integration of SAM-based semantic augmentation and homography-based smoothness for enhanced unsupervised optical flow estimation.

Findings

Outperforms state-of-the-art on KITTI and Sintel datasets

Produces sharp object boundaries in optical flow

Generalizes well across different domains

Abstract

Traditional unsupervised optical flow methods are vulnerable to occlusions and motion boundaries due to lack of object-level information. Therefore, we propose UnSAMFlow, an unsupervised flow network that also leverages object information from the latest foundation model Segment Anything Model (SAM). We first include a self-supervised semantic augmentation module tailored to SAM masks. We also analyze the poor gradient landscapes of traditional smoothness losses and propose a new smoothness definition based on homography instead. A simple yet effective mask feature module has also been added to further aggregate features on the object level. With all these adaptations, our method produces clear optical flow estimation with sharp boundaries around objects, which outperforms state-of-the-art methods on both KITTI and Sintel datasets. Our method also generalizes well across domains and runs very efficiently.