Unified Mask Embedding and Correspondence Learning for Self-Supervised Video Segmentation

TL;DR

This paper introduces a self-supervised video object segmentation method that learns to perform mask-guided segmentation by clustering pixels and embedding object context, achieving state-of-the-art results without labeled data.

Contribution

It presents a unified framework combining dense correspondence and object-level embedding for self-supervised video segmentation, improving over previous methods that rely on pixel correlation.

Findings

Sets new state-of-the-art on DAVIS17 and YouTube-VOS benchmarks.

Narrowing the performance gap between self-supervised and fully-supervised VOS.

Effective pseudo-label generation through pixel clustering.

Abstract

The objective of this paper is self-supervised learning of video object segmentation. We develop a unified framework which simultaneously models cross-frame dense correspondence for locally discriminative feature learning and embeds object-level context for target-mask decoding. As a result, it is able to directly learn to perform mask-guided sequential segmentation from unlabeled videos, in contrast to previous efforts usually relying on an oblique solution - cheaply "copying" labels according to pixel-wise correlations. Concretely, our algorithm alternates between i) clustering video pixels for creating pseudo segmentation labels ex nihilo; and ii) utilizing the pseudo labels to learn mask encoding and decoding for VOS. Unsupervised correspondence learning is further incorporated into this self-taught, mask embedding scheme, so as to ensure the generic nature of the learnt representation and avoid cluster degeneracy. Our algorithm sets state-of-the-arts on two standard benchmarks (i.e., DAVIS17 and YouTube-VOS), narrowing the gap between self- and fully-supervised VOS, in terms of both performance and network architecture design.