UVOSAM: A Mask-free Paradigm for Unsupervised Video Object Segmentation via Segment Anything Model

TL;DR

This paper introduces UVOSAM, a novel mask-free approach for unsupervised video object segmentation that leverages the Segment Anything Model and a specialized tracker to outperform existing methods without requiring mask annotations.

Contribution

UVOSAM is the first mask-free UVOS method utilizing SAM with a new tracker and attention mechanism, achieving superior results over mask-supervised approaches.

Findings

Outperforms existing mask-supervised UVOS methods on DAVIS2017-unsupervised and YoutubeVIS datasets.

Demonstrates strong generalization to weakly-annotated video datasets.

Uses a novel spatial-temporal decoupled deformable attention mechanism.

Abstract

The current state-of-the-art methods for unsupervised video object segmentation (UVOS) require extensive training on video datasets with mask annotations, limiting their effectiveness in handling challenging scenarios. However, the Segment Anything Model (SAM) introduces a new prompt-driven paradigm for image segmentation, offering new possibilities. In this study, we investigate SAM's potential for UVOS through different prompt strategies. We then propose UVOSAM, a mask-free paradigm for UVOS that utilizes the STD-Net tracker. STD-Net incorporates a spatial-temporal decoupled deformable attention mechanism to establish an effective correlation between intra- and inter-frame features, remarkably enhancing the quality of box prompts in complex video scenes. Extensive experiments on the DAVIS2017-unsupervised and YoutubeVIS19\&21 datasets demonstrate the superior performance of UVOSAM without mask supervision compared to existing mask-supervised methods, as well as its ability to generalize to weakly-annotated video datasets. Code can be found at https://github.com/alibaba/UVOSAM.