Hierarchical Visual Feature Fusion for Robust Out‐of‐Distribution Detection under Semantic and Image‐Level Covariate Shifts
Computer Graphics Forum · 13.11.2026
DOI: 10.1111/cgf.70619 ↗Kurz gefasst
Fachartikel in der Zeitschrift Computer Graphics Forum, online erschienen am 13.11.2026. Verfasst von Chao Hou und 7 weiteren Autor:innen. Die Arbeit nennt 75 Literaturangaben. Gefördert von: Graduate School, Chongqing University, National Natural Science Foundation of China.
Abstract (Originalsprache)
Deep neural networks (DNNs) have become a fundamental component of modern visual computing systems, yet their reliability often degrades on out‐of‐distribution (OOD) inputs, i.e., test images that deviate from the training distribution. Such deviations may arise from high‐level semantic shifts, where images belong to unseen categories, or from image‐level covariate shifts, such as noise, blur, illumination changes, and compression artifacts. Although covariate shifts often preserve the original class identity, they can substantially distort low‐level visual statistics and alter model responses, making them difficult to detect with methods designed primarily for semantic novelty. This challenge becomes more severe when semantic and covariate shifts coexist, as they jointly affect different levels of visual abstraction. To address this issue, we propose AFDisc, a unified OOD detection framework that exploits the hierarchical representations of DNNs. AFDisc is motivated by the observation that shallow features are more sensitive to appearance variations, while deep features encode richer semantic information. Based on this complementarity, we introduce an attentive fusion module that adaptively integrates appearance‐sensitive cues from shallow layers and semantic cues from deep layers into a unified representation for detecting diverse distribution shifts. We further construct auxiliary supervision signals through controlled appearance perturbations, enabling the model to better capture the spectrum of real‐world visual degradations. Extensive experiments under semantic and covariate shift settings demonstrate AFDisc's effectiveness and its superior performance over prior state‐of‐the‐art methods.
Förderung
- Graduate School, Chongqing University
- National Natural Science Foundation of China
Zitieren
APA
Hou, C., Ma, C., Du, Z., Lv, X., Tang, K., Zhu, B., He, S., & Xiang, T. (2026). Hierarchical Visual Feature Fusion for Robust Out‐of‐Distribution Detection under Semantic and Image‐Level Covariate Shifts. Computer Graphics Forum. https://doi.org/10.1111/cgf.70619
Vancouver
Hou C, Ma C, Du Z, Lv X, Tang K, Zhu B, et al. Hierarchical Visual Feature Fusion for Robust Out‐of‐Distribution Detection under Semantic and Image‐Level Covariate Shifts. Computer Graphics Forum. 2026. doi:10.1111/cgf.70619
BibTeX
@article{Hou2026,
author = {Hou, Chao and Ma, Chuan and Du, Ziyong and Lv, Xiao and Tang, Keke and Zhu, Bin and He, Shengfeng and Xiang, Tao},
title = {Hierarchical Visual Feature Fusion for Robust Out‐of‐Distribution Detection under Semantic and Image‐Level Covariate Shifts},
journal = {Computer Graphics Forum},
year = {2026},
doi = {10.1111/cgf.70619}
}
EUVIMED