Self-supervised video object segmentation via pseudo label rectification

Open Webpage

Pinxue Guo, Wei Zhang, Xiaoqiang Li, Jianping Fan, Wenqiang Zhang

Published: 2025, Last Modified: 24 Oct 2025Pattern Recognit. 2025EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•Propose a novel self-supervised framework with bi-decoders for VOS.•Reduce the gap between pseudo labels and unknown ground-truth for model learning.•Pseudo label rectification via bi-decoders mutual learning.•Achieve SOTA performance on the challenging benchmark datasets.