Counterfactual Co-occurring Learning for Bias Mitigation in Weakly-supervised Object Localization

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Shao, Feifei, Luo, Yawei, Chen, Lei, Liu, Ping, Yang, Wei, Yang, Yi, Xiao, Jun
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913259369529344
author Shao, Feifei
Luo, Yawei
Chen, Lei
Liu, Ping
Yang, Wei
Yang, Yi
Xiao, Jun
author_facet Shao, Feifei
Luo, Yawei
Chen, Lei
Liu, Ping
Yang, Wei
Yang, Yi
Xiao, Jun
contents Contemporary weakly-supervised object localization (WSOL) methods have primarily focused on addressing the challenge of localizing the most discriminative region while largely overlooking the relatively less explored issue of biased activation -- incorrectly spotlighting co-occurring background with the foreground feature. In this paper, we conduct a thorough causal analysis to investigate the origins of biased activation. Based on our analysis, we attribute this phenomenon to the presence of co-occurring background confounders. Building upon this profound insight, we introduce a pioneering paradigm known as Counterfactual Co-occurring Learning (CCL), meticulously engendering counterfactual representations by adeptly disentangling the foreground from the co-occurring background elements. Furthermore, we propose an innovative network architecture known as Counterfactual-CAM. This architecture seamlessly incorporates a perturbation mechanism for counterfactual representations into the vanilla CAM-based model. By training the WSOL model with these perturbed representations, we guide the model to prioritize the consistent foreground content while concurrently reducing the influence of distracting co-occurring backgrounds. To the best of our knowledge, this study represents the initial exploration of this research direction. Our extensive experiments conducted across multiple benchmarks validate the effectiveness of the proposed Counterfactual-CAM in mitigating biased activation.
format Preprint
id arxiv_https___arxiv_org_abs_2305_15354
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Counterfactual Co-occurring Learning for Bias Mitigation in Weakly-supervised Object Localization
Shao, Feifei
Luo, Yawei
Chen, Lei
Liu, Ping
Yang, Wei
Yang, Yi
Xiao, Jun
Computer Vision and Pattern Recognition
Contemporary weakly-supervised object localization (WSOL) methods have primarily focused on addressing the challenge of localizing the most discriminative region while largely overlooking the relatively less explored issue of biased activation -- incorrectly spotlighting co-occurring background with the foreground feature. In this paper, we conduct a thorough causal analysis to investigate the origins of biased activation. Based on our analysis, we attribute this phenomenon to the presence of co-occurring background confounders. Building upon this profound insight, we introduce a pioneering paradigm known as Counterfactual Co-occurring Learning (CCL), meticulously engendering counterfactual representations by adeptly disentangling the foreground from the co-occurring background elements. Furthermore, we propose an innovative network architecture known as Counterfactual-CAM. This architecture seamlessly incorporates a perturbation mechanism for counterfactual representations into the vanilla CAM-based model. By training the WSOL model with these perturbed representations, we guide the model to prioritize the consistent foreground content while concurrently reducing the influence of distracting co-occurring backgrounds. To the best of our knowledge, this study represents the initial exploration of this research direction. Our extensive experiments conducted across multiple benchmarks validate the effectiveness of the proposed Counterfactual-CAM in mitigating biased activation.
title Counterfactual Co-occurring Learning for Bias Mitigation in Weakly-supervised Object Localization
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2305.15354