HQD-EM: Robust VQA Through Hierarchical Question Decomposition Bias Module and Ensemble Adaptive Angular Margin Loss

SeongHyeon Noh and Jae Won Cho ()
Additional contact information
SeongHyeon Noh: Department of Computer Engineering, Sejong University, Gwangjin-gu, Seoul 05006, Republic of Korea
Jae Won Cho: Department of Artificial Intelligence & Data Science, Sejong University, Gwangjin-gu, Seoul 05006, Republic of Korea

Mathematics, 2025, vol. 13, issue 22, 1-19

Abstract: Recent studies in Visual Question Answering (VQA) have revealed that models often rely heavily on language priors rather than vision–language understanding, leading to poor generalization under distribution shifts. To address this challenge, we propose HQD-EM, a unified debiasing framework that combines the Hierarchical Question Decomposition (HQD) module with an Ensemble adaptive angular Margin (EM) loss. HQD systematically decomposes questions into multi-granular representations to capture layered language biases, while EM leverages bias confidence to modulate per-sample decision margins dynamically. Our method integrates an ensemble-based method with adaptive margin learning in an end-to-end trainable architecture. Experiments on VQA benchmarks demonstrate that HQD-EM significantly outperforms prior works on VQA-CP2 and VQA-CP1.

Keywords: Visual Question Answering; vision and language; bias issues; robustness (search for similar items in EconPapers)
JEL-codes: C (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/2227-7390/13/22/3656/pdf (application/pdf)
https://www.mdpi.com/2227-7390/13/22/3656/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jmathe:v:13:y:2025:i:22:p:3656-:d:1794774

Access Statistics for this article

Mathematics is currently edited by Ms. Emma He

More articles in Mathematics from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().