Segment Anything Model (SAM) Enhanced Pseudo Labels for Weakly Supervised Semantic Segmentation

TL;DR

This paper introduces a novel approach that combines Class Activation Maps with the Segment Anything Model to generate high-quality pseudo-labels, significantly improving weakly supervised semantic segmentation performance.

Contribution

It presents a simple, versatile method leveraging SAM masks with CAM pseudo-labels to produce more accurate object-aware labels for WSSS, compatible with existing methods.

Findings

Consistent improvements over state-of-the-art on PASCAL VOC and MS-COCO.

Enhanced pseudo-label quality leads to better segmentation accuracy.

Method is easy to integrate into existing WSSS frameworks.

Abstract

Weakly supervised semantic segmentation (WSSS) aims to bypass the need for laborious pixel-level annotation by using only image-level annotation. Most existing methods rely on Class Activation Maps (CAM) to derive pixel-level pseudo-labels and use them to train a fully supervised semantic segmentation model. Although these pseudo-labels are class-aware, indicating the coarse regions for particular classes, they are not object-aware and fail to delineate accurate object boundaries. To address this, we introduce a simple yet effective method harnessing the Segment Anything Model (SAM), a class-agnostic foundation model capable of producing fine-grained instance masks of objects, parts, and subparts. We use CAM pseudo-labels as cues to select and combine SAM masks, resulting in high-quality pseudo-labels that are both class-aware and object-aware. Our approach is highly versatile and can be easily integrated into existing WSSS methods without any modification. Despite its simplicity, our approach shows consistent gain over the state-of-the-art WSSS methods on both PASCAL VOC and MS-COCO datasets.