Have a personal or library account? Click to login
Inverse Meets Distillation: Heterogeneous Teacher–Assistant Dual-Path Learning for Unsupervised Defect Detection Cover

Inverse Meets Distillation: Heterogeneous Teacher–Assistant Dual-Path Learning for Unsupervised Defect Detection

Open Access
|Feb 2026

Abstract

Unsupervised defect detection is crucial for industrial inspection, but teacher–student (T–S) frameworks tend to overfit a single teacher’s feature manifold, leading to poor generalization on subtle anomalies. We introduce TAD++, a dual-path distillation framework that combines heterogeneous Teacher–Assistant–Student (T–A–S) guidance with a pseudo-defect inverse-distillation branch. A compact assistant, structurally distinct from the teacher, is trained to co-distill the student, thereby mitigating single-teacher bias. In parallel, the inverse-distillation path tasks the student with reconstructing normal appearances from defect-injected inputs, serving as a regularization term to prevent anomaly leakage. A dynamic attention weighting module adaptively fuses these heterogeneous guidance signals. Crucially, the assistant, inverse branch, and weight modules are strictly training-only. This design ensures that while TAD++ benefits from a rigorous multi-phase optimization, it maintains zero additional inference latency and memory overhead compared to standard T–S baselines. On MVTec AD, BTAD, and VisA, TAD++ achieves consistent improvements in both image-level detection and pixel-level localization, with extensive ablations confirming the efficacy of the heterogeneous dual-path design.

Language: English
Page range: 275 - 291
Submitted on: Sep 26, 2025
|
Accepted on: Jan 29, 2026
|
Published on: Feb 25, 2026
Published by: SAN University
In partnership with: Paradigm Publishing Services
Publication frequency: 4 issues per year

© 2026 Xin Zhan, Yaqian Li, Ruihao Chen, Wenming Zhang, Haibin Li, published by SAN University
This work is licensed under the Creative Commons Attribution 4.0 License.