Weak-shot Semantic Segmentation by Transferring Semantic Affinity and Boundary

TL;DR

This paper introduces a weak-shot semantic segmentation approach that leverages fully-annotated base categories to improve segmentation of novel categories with only image-level labels, by transferring semantic affinity and boundary information.

Contribution

It proposes a novel method to transfer semantic affinity and boundary from base to novel categories, enhancing CAM quality in weakly-supervised segmentation.

Findings

Significant performance improvement on PASCAL VOC 2012 dataset

Effective transfer of semantic affinity and boundary improves CAMs

Outperforms existing WSSS methods on novel categories

Abstract

Weakly-supervised semantic segmentation (WSSS) with image-level labels has been widely studied to relieve the annotation burden of the traditional segmentation task. In this paper, we show that existing fully-annotated base categories can help segment objects of novel categories with only image-level labels, even if base categories and novel categories have no overlap. We refer to this task as weak-shot semantic segmentation, which could also be treated as WSSS with auxiliary fully-annotated categories. Recent advanced WSSS methods usually obtain class activation maps (CAMs) and refine them by affinity propagation. Based on the observation that semantic affinity and boundary are class-agnostic, we propose a method under the WSSS framework to transfer semantic affinity and boundary from base to novel categories. As a result, we find that pixel-level annotation of base categories can facilitate affinity learning and propagation, leading to higher-quality CAMs of novel categories. Extensive experiments on PASCAL VOC 2012 dataset prove that our method significantly outperforms WSSS baselines on novel categories.