Region-CAM: Towards Accurate Object Regions in Class Activation Maps for Weakly Supervised Learning Tasks

TL;DR

Region-CAM introduces a novel method for generating more accurate and boundary-aligned object activation maps in weakly supervised learning, significantly improving segmentation and localization performance over traditional CAM methods.

Contribution

The paper proposes Region-CAM, a new activation method that enhances object region coverage and boundary precision in CAMs by using semantic information propagation considering gradients and features.

Findings

Achieves 60.12% mIoU on PASCAL VOC training set, 13.61% improvement over CAM.

Improves MS COCO validation mIoU to 36.38%, 16.23% higher than CAM.

Attains 51.7% Top-1 Localization accuracy, outperforming LayerCAM by 4.5%.

Abstract

Class Activation Mapping (CAM) methods are widely applied in weakly supervised learning tasks due to their ability to highlight object regions. However, conventional CAM methods highlight only the most discriminative regions of the target. These highlighted regions often fail to cover the entire object and are frequently misaligned with object boundaries, thereby limiting the performance of downstream weakly supervised learning tasks, particularly Weakly Supervised Semantic Segmentation (WSSS), which demands pixel-wise accurate activation maps to get the best results. To alleviate the above problems, we propose a novel activation method, Region-CAM. Distinct from network feature weighting approaches, Region-CAM generates activation maps by extracting semantic information maps (SIMs) and performing semantic information propagation (SIP) by considering both gradients and features in each of the stages of the baseline classification model. Our approach highlights a greater proportion of object regions while ensuring activation maps to have precise boundaries that align closely with object edges. Region-CAM achieves 60.12% and 58.43% mean intersection over union (mIoU) using the baseline model on the PASCAL VOC training and validation datasets, respectively, which are improvements of 13.61% and 13.13% over the original CAM (46.51% and 45.30%). On the MS COCO validation set, Region-CAM achieves 36.38%, a 16.23% improvement over the original CAM (20.15%). We also demonstrate the superiority of Region-CAM in object localization tasks, using the ILSVRC2012 validation set. Region-CAM achieves 51.7% in Top-1 Localization accuracy Loc1. Compared with LayerCAM, an activation method designed for weakly supervised object localization, Region-CAM achieves 4.5% better performance in Loc1.