STAN: Spatio-Temporal Adversarial Networks for Abnormal Event Detection

TL;DR

This paper introduces STAN, a spatio-temporal adversarial network for abnormal event detection that synthesizes and discriminates normal patterns to identify anomalies in video sequences.

Contribution

The paper presents a novel spatio-temporal adversarial network architecture with a generator and discriminator trained to detect abnormal events in videos.

Findings

Achieved competitive performance with state-of-the-art methods.

Enabled visualization of abnormal event locations.

Effectively encodes spatio-temporal features of normal patterns.

Abstract

In this paper, we propose a novel abnormal event detection method with spatio-temporal adversarial networks (STAN). We devise a spatio-temporal generator which synthesizes an inter-frame by considering spatio-temporal characteristics with bidirectional ConvLSTM. A proposed spatio-temporal discriminator determines whether an input sequence is real-normal or not with 3D convolutional layers. These two networks are trained in an adversarial way to effectively encode spatio-temporal features of normal patterns. After the learning, the generator and the discriminator can be independently used as detectors, and deviations from the learned normal patterns are detected as abnormalities. Experimental results show that the proposed method achieved competitive performance compared to the state-of-the-art methods. Further, for the interpretation, we visualize the location of abnormal events detected by the proposed networks using a generator loss and discriminator gradients.