AI Paper Atlas

Research map / Label-efficient and robust learning / OOD and anomaly detection

Multiresolution Knowledge Distillation for Anomaly Detection

Mohammadreza Salehi, Niousha Sadjadi, Soroosh Baselizadeh, Mohammad H. Rohban, Hamid R. Rabiee

CVPR 2021 · Label-efficient and robust learning › OOD and anomaly detection · approach: Unified multi-class anomaly detection

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 changedWhereMeasuredEffect
Add additional intermediate max-pooling-layer hints to distillation (ablation)Multiresolution Knowledge Distillation · MVTecADAUROC: 79.22 -> 83.02 -> 87.74Improved
Add additional intermediate max-pooling-layer hints to distillation (ablation)Multiresolution Knowledge Distillation · MNISTAUROC: 95.33 -> 97.74 -> 98.71Improved
Use a smaller cloner instead of a source-sized cloner (ablation)Multiresolution Knowledge Distillation · MVTecADAUROC: 85.58 -> 87.74Improved
Use a smaller cloner instead of a source-sized cloner (ablation)Multiresolution Knowledge Distillation · CIFAR-10AUROC: 86.87 -> 87.18Improved
Add directional loss to MSE loss (ablation)Multiresolution Knowledge Distillation · CIFAR-10AUROC: 70.16 -> 87.2Improved
Add MSE loss to directional loss (ablation)Multiresolution Knowledge Distillation · MVTecADAUROC: 83.74 -> 87.74Mixed
Add 20 degree rotation and scaling in range [0.9, 1.05] (ablation)Multiresolution Knowledge Distillation · HeadCTAUROC: 78.04 ± 0.225 -> 80.42 ± 0.006Improved
Add Gaussian filtering to gradient localization maps (ablation)Multiresolution Knowledge Distillation · MVTecADpixel-wise AUROC: 86.16 -> 90.51Improved
Use SmoothGrad rather than simple gradients without Gaussian filtering (ablation)Multiresolution Knowledge Distillation · MVTecADpixel-wise AUROC: 86.16 -> 86.97Improved
Use SmoothGrad rather than simple gradients with Gaussian filtering (ablation)Multiresolution Knowledge Distillation · MVTecADpixel-wise AUROC: 90.51 -> 90.54No clear effect
Use GBP rather than simple gradients without Gaussian filtering (ablation)Multiresolution Knowledge Distillation · MVTecADpixel-wise AUROC: 86.16 -> 84.38Hurt

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy