Betrayed by Attention: A Simple yet Effective Approach for Self-supervised Video Object Segmentation

TL;DR

This paper introduces a simple, effective self-supervised video object segmentation method leveraging DINO-pretrained Transformers' structural dependencies, achieving state-of-the-art results without auxiliary modalities or complex attention mechanisms.

Contribution

The authors propose a novel approach that uses a single spatio-temporal Transformer and hierarchical clustering on DINO features for self-supervised VOS, simplifying previous methods.

Findings

Achieves state-of-the-art results on multiple benchmarks.

Excels in complex multi-object video segmentation tasks.

Operates without auxiliary modalities or iterative slot attention.

Abstract

In this paper, we propose a simple yet effective approach for self-supervised video object segmentation (VOS). Our key insight is that the inherent structural dependencies present in DINO-pretrained Transformers can be leveraged to establish robust spatio-temporal correspondences in videos. Furthermore, simple clustering on this correspondence cue is sufficient to yield competitive segmentation results. Previous self-supervised VOS techniques majorly resort to auxiliary modalities or utilize iterative slot attention to assist in object discovery, which restricts their general applicability and imposes higher computational requirements. To deal with these challenges, we develop a simplified architecture that capitalizes on the emerging objectness from DINO-pretrained Transformers, bypassing the need for additional modalities or slot attention. Specifically, we first introduce a single spatio-temporal Transformer block to process the frame-wise DINO features and establish spatio-temporal dependencies in the form of self-attention. Subsequently, utilizing these attention maps, we implement hierarchical clustering to generate object segmentation masks. To train the spatio-temporal block in a fully self-supervised manner, we employ semantic and dynamic motion consistency coupled with entropy normalization. Our method demonstrates state-of-the-art performance across multiple unsupervised VOS benchmarks and particularly excels in complex real-world multi-object video segmentation tasks such as DAVIS-17-Unsupervised and YouTube-VIS-19. The code and model checkpoints will be released at https://github.com/shvdiwnkozbw/SSL-UVOS.