Pseudo-label Alignment for Semi-supervised Instance Segmentation

TL;DR

This paper introduces PAIS, a novel semi-supervised instance segmentation framework that uses dynamic loss weighting to effectively utilize pseudo-labels, significantly improving performance especially with limited labeled data.

Contribution

The paper proposes a new pseudo-label aligning framework with a dynamic loss adjustment mechanism, enhancing semi-supervised instance segmentation performance.

Findings

PAIS achieves 21.2 mAP with 1% labeled data on COCO.

PAIS outperforms the state-of-the-art NoisyBoundary by over 12 points.

The dynamic aligning loss effectively leverages pseudo-labels with varying quality.

Abstract

Pseudo-labeling is significant for semi-supervised instance segmentation, which generates instance masks and classes from unannotated images for subsequent training. However, in existing pipelines, pseudo-labels that contain valuable information may be directly filtered out due to mismatches in class and mask quality. To address this issue, we propose a novel framework, called pseudo-label aligning instance segmentation (PAIS), in this paper. In PAIS, we devise a dynamic aligning loss (DALoss) that adjusts the weights of semi-supervised loss terms with varying class and mask score pairs. Through extensive experiments conducted on the COCO and Cityscapes datasets, we demonstrate that PAIS is a promising framework for semi-supervised instance segmentation, particularly in cases where labeled data is severely limited. Notably, with just 1\% labeled data, PAIS achieves 21.2 mAP (based on Mask-RCNN) and 19.9 mAP (based on K-Net) on the COCO dataset, outperforming the current state-of-the-art model, \ie, NoisyBoundary with 7.7 mAP, by a margin of over 12 points. Code is available at: \url{https://github.com/hujiecpp/PAIS}.