Heuristic Hyperparameter Choice for Image Anomaly Detection - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Heuristic Hyperparameter Choice for Image Anomaly Detection

Résumé

Anomaly detection (AD) in images is a fundamental computer vision problem by deep learning neural network to identify images deviating significantly from normality. The deep features extracted from pre-trained models have been proved to be essential for AD based on multivariate Gaussian distribution analysis. However, since models are usually pre-trained on a large dataset for classification tasks such as ImageNet, they might produce lots of redundant features for AD, which increases computational cost and degrades the performance. To reduce the number of dimensions of these features, we apply Negated Principal Component Analysis (NPCA) combined with heuristics to optimize the number of its hyperparameter to get as few features as possible while ensuring a good performance.

Dates et versions

hal-04319319 , version 1 (02-12-2023)

Licence

Paternité - Partage selon les Conditions Initiales

Identifiants

Citer

Zeyu Jiang, Joao P C Bertoldo, Etienne Decencière. Heuristic Hyperparameter Choice for Image Anomaly Detection. 2023 Twelfth International Conference on Image Processing Theory, Tools and Applications (IPTA), Oct 2023, Paris, France. pp.1-5, ⟨10.1109/IPTA59101.2023.10320035⟩. ⟨hal-04319319⟩
23 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More