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.