Combinational Class Activation Maps for Weakly Supervised Object Localization

TL;DR

This paper introduces NL-CCAM, a novel approach for weakly supervised object localization that combines multiple class activation maps and incorporates non-local modules to improve localization accuracy and generalization.

Contribution

The paper proposes combinational class activation maps (CCAM) and integrates non-local modules, enhancing object localization by considering multiple class activations and spatial relationships.

Findings

NL-CCAM outperforms previous methods on ILSVRC 2016 and CUB-200-2011 benchmarks.

The method effectively suppresses background regions and highlights foreground objects.

Demonstrates strong generalization capabilities across datasets.

Abstract

Weakly supervised object localization has recently attracted attention since it aims to identify both class labels and locations of objects by using image-level labels. Most previous methods utilize the activation map corresponding to the highest activation source. Exploiting only one activation map of the highest probability class is often biased into limited regions or sometimes even highlights background regions. To resolve these limitations, we propose to use activation maps, named combinational class activation maps (CCAM), which are linear combinations of activation maps from the highest to the lowest probability class. By using CCAM for localization, we suppress background regions to help highlighting foreground objects more accurately. In addition, we design the network architecture to consider spatial relationships for localizing relevant object regions. Specifically, we integrate non-local modules into an existing base network at both low- and high-level layers. Our final model, named non-local combinational class activation maps (NL-CCAM), obtains superior performance compared to previous methods on representative object localization benchmarks including ILSVRC 2016 and CUB-200-2011. Furthermore, we show that the proposed method has a great capability of generalization by visualizing other datasets.