Robust Anomaly Detection in Videos Using Multilevel Representations

Authors

  • Hung Vu Deakin University
  • Tu Dinh Nguyen Monash University
  • Trung Le Monash University
  • Wei Luo Deakin University
  • Dinh Phung Monash University

DOI:

https://doi.org/10.1609/aaai.v33i01.33015216

Abstract

Detecting anomalies in surveillance videos has long been an important but unsolved problem. In particular, many existing solutions are overly sensitive to (often ephemeral) visual artifacts in the raw video data, resulting in false positives and fragmented detection regions. To overcome such sensitivity and to capture true anomalies with semantic significance, one natural idea is to seek validation from abstract representations of the videos. This paper introduces a framework of robust anomaly detection using multilevel representations of both intensity and motion data. The framework consists of three main components: 1) representation learning using Denoising Autoencoders, 2) level-wise representation generation using Conditional Generative Adversarial Networks, and 3) consolidating anomalous regions detected at each representation level. Our proposed multilevel detector shows a significant improvement in pixel-level Equal Error Rate, namely 11.35%, 12.32% and 4.31% improvement in UCSD Ped 1, UCSD Ped 2 and Avenue datasets respectively. In addition, the model allowed us to detect mislabeled anomalies in the UCDS Ped 1.

Downloads

Published

2019-07-17

How to Cite

Vu, H., Nguyen, T. D., Le, T., Luo, W., & Phung, D. (2019). Robust Anomaly Detection in Videos Using Multilevel Representations. Proceedings of the AAAI Conference on Artificial Intelligence, 33(01), 5216-5223. https://doi.org/10.1609/aaai.v33i01.33015216

Issue

Section

AAAI Technical Track: Machine Learning