Improving Weakly-supervised Object Localization via Causal Intervention

TL;DR

This paper introduces CI-CAM, a causal intervention approach that improves weakly-supervised object localization by addressing co-occurrence confounders, leading to more accurate object boundary detection in images.

Contribution

The paper proposes a novel causal intervention method, CI-CAM, to eliminate confounding effects in weakly-supervised object localization, enhancing localization accuracy.

Findings

Significant improvement on CUB-200-2011 dataset (58.39% vs 52.4%)

Performs comparably to state-of-the-art on ImageNet

Effectively reduces co-occurrence confounder influence

Abstract

The recent emerged weakly supervised object localization (WSOL) methods can learn to localize an object in the image only using image-level labels. Previous works endeavor to perceive the interval objects from the small and sparse discriminative attention map, yet ignoring the co-occurrence confounder (e.g., bird and sky), which makes the model inspection (e.g., CAM) hard to distinguish between the object and context. In this paper, we make an early attempt to tackle this challenge via causal intervention (CI). Our proposed method, dubbed CI-CAM, explores the causalities among images, contexts, and categories to eliminate the biased co-occurrence in the class activation maps thus improving the accuracy of object localization. Extensive experiments on several benchmarks demonstrate the effectiveness of CI-CAM in learning the clear object boundaries from confounding contexts. Particularly, in CUB-200-2011 which severely suffers from the co-occurrence confounder, CI-CAM significantly outperforms the traditional CAM-based baseline (58.39% vs 52.4% in top-1 localization accuracy). While in more general scenarios such as ImageNet, CI-CAM can also perform on par with the state of the arts.