HiProtoAD: Feature Reconstruction Guided by Cluster-Separated Hierarchical Prototypes for Multi-Class Unsupervised Anomaly Detection

Prevailing Transformer-based feature reconstruction methods for multi-class unsupervised visual anomaly detection suffer from identity mapping, while the underlying causes remain insufficiently understood. Through a systematic analysis, we reveal that identity mapping is closely related to anomaly i...

ver descrição completa

Detalhes bibliográficos

Principais autores:	Chaoming Li, Guoqing Yu, Wenwen Ye
Formato:	Artigo
Idioma:	Inglês
Publicado em:	IEEE 2026-01-01
coleção:	IEEE Access
Assuntos:	Visual anomaly detection prototype learning feature reconstruction vision transformer cross-attention
Acesso em linha:	https://ieeexplore.ieee.org/document/11585790/
Tags:	Adicionar Tag Sem tags, seja o primeiro a adicionar uma tag!