Anomaly Detection in Image and Video with Improved False Positive Rate

Please enable javascript in your browser.

Salehi Dehnavi, Mohammad Reza | 2020

664 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 53716 (19)
University: Sharif University of Technology
Department: Computer Engineering
Advisor(s): Rabiee, Hamid Reza; Rohban, Mohammad Reza
Abstract:
Autoencoder, as an essential part of many anomaly detection methods, is lacking flexibility on normal data in complex datasets. U-Net is proved to be effective for this purpose but overfits on the training data if trained by just using reconstruction error similar to other AE-based frameworks. Puzzle-solving, as a pretext task of self-supervised learning (SSL) methods, has earlier proved its ability in learning semantically meaningful features. We show that training U-Nets based on this task is an effective remedy that prevents overfitting and facilitates learning beyond pixel-level features. Shortcut solutions, however,are a big challenge in SSL tasks, including jigsaw puzzles. We propose adversarial robust training as an effective automatic shortcut removal. We achieve competitive or superior results compared to the State of the Art (SOTA)anomaly detection methods on various toy and real-world datasets. Unlike many competitors, the proposed framework is stable, fast, data-efficient, and does not require unprincipled early stopping
Keywords:
Autoencoder ; Adversarial Training ; Self-Supervised Learning ; Novelty Detection ; One-Class Learning ; Adversarial Robust Training