Research map / Adversarial robustness and security / Attacks on text and graphs
ABIGX: A Unified Framework for Explainable Fault Detection and Classification
Yue Zhuo, Jinchuan Qian, Junhua Zheng, Xiaoyu Jiang, Zhihuan Song, Duxin Chen, Wenwu Yu, Zhiqiang Ge
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add the classification SPE index for adversarial fault reconstruction (instead of minimizing normal-class softmax confidence) (ablation) | ABIGX · TEP NN classifier | Consistency-ADD: 0.726 -> 0.890 | Improved |
| add the classification SPE index (ablation) | ABIGX · TEP NN classifier | Consistency-DEL: 0.135 -> 0.032 | Improved |
| add the classification SPE index (ablation) | ABIGX · TEP NN classifier | Correctness-AUC: 0.608 -> 0.557 | Hurt |
| use AFR-reconstructed baseline instead of normality-sample baseline of IG | IG · TEP AE detector | Consistency-ADD: 0.913 -> 0.922 | Improved |
| use AFR-reconstructed baseline instead of IG baseline | IG · TEP NN classifier | Consistency-DEL: 0.043 -> 0.032 | Improved |
| use AFR-reconstructed baseline instead of IG baseline | IG · TEP NN classifier | Correctness-AUC: 0.415 -> 0.557 | Improved |
| use ABIGX instead of IG on wafer maps | IG · WM-811K CNN | Correctness-AUC: 0.814 -> 0.839 | Improved |
| use ABIGX instead of saliency map on wafer maps | saliency map · WM-811K CNN | Consistency-ADD: 0.623 -> 0.859 | Improved |
| use one-variable-at-a-time reconstruction (OneVar) instead of joint reconstruction | ABIGX · TEP AE detector | Correctness-AUC: 0.382 -> 0.478 | Improved |
| use one-variable-at-a-time reconstruction (OneVar) instead of joint reconstruction | ABIGX · TEP NN classifier | Consistency-ADD: 0.890 -> 0.723 | Hurt |
| use MILP instead of PGD for AFR when comparing ABIGX-OneVar with RBC (ablation) | ABIGX-OneVar · TEP PCA detector | mean difference to RBC: 8.87e-3 -> 3.5e-5 | Improved |
| use MILP instead of PGD for AFR when comparing ABIGX with CP (ablation) | ABIGX · TEP PCA detector | mean difference to CP: 1.62e-6 -> 1.25e-6 | Improved |
Closest papers by meaning
- Long-Tailed Anomaly Detection with Learnable Class Names (CVPR 2024 · similarity 0.81)
- AR-Pro: Counterfactual Explanations for Anomaly Repair with Formal Properties (NeurIPS 2024 · similarity 0.81)
- A Theory of Fault-Tolerant Learning (ICML 2024 · similarity 0.81)
- On Gradient-like Explanation under a Black-box Setting: When Black-box Explanations Become as Good as White-box (ICML 2024 · similarity 0.80)
- AnomalyNCD: Towards Novel Anomaly Class Discovery in Industrial Scenarios (CVPR 2025 · similarity 0.80)
- ResAD: A Simple Framework for Class Generalizable Anomaly Detection (NeurIPS 2024 · similarity 0.80)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
