Search

The interest in video anomaly detection systems that can detect different types of anomalies, such as violent behaviours in surveillance videos, has gained traction in recent years. The current approaches employ deep learning to perform anomaly detection in videos, but this approach has multiple problems. For example, deep learning in general has issues with noise, concept drift, explainability, and training data volumes. Additionally, anomaly detection in itself is a complex task and faces challenges such as unknownness, heterogeneity, and class imbalance. Anomaly detection using deep learning is therefore mainly constrained to generative models such as generative adversarial networks and autoencoders due to their unsupervised nature; however, even they suffer from general deep learning issues and are hard to properly train. In this paper, we explore the capabilities of the Hierarchical Temporal Memory (HTM) algorithm to perform anomaly detection in videos, as it has favorable properties such as noise tolerance and online learning which combats concept drift. We introduce a novel version of HTM, named GridHTM, which is a grid-based HTM architecture specifically for anomaly detection in complex videos such as surveillance footage. We have tested GridHTM using the VIRAT video surveillance dataset, and the subsequent evaluation results and online learning capabilities prove the great potential of using our system for real-time unsupervised anomaly detection in complex videos.

This paper introduces a two-stage approach to the detection of people eating and/or drinking for the purposes of surveillance of daily life. With the sole use of wearable accelerometer sensor attached to somebody’s (man or a woman) wrists, this two-stage approach consists of feature extraction followed by classification. At the first stage, based on the limb’s three dimensional kinematics movement model and the Extended Kalman Filter (EKF), the realtime arm movement features described by Euler angles are extracted from the raw accelerometer measurement data. In the latter stage, the Hierarchical Temporal Memory (HTM) network is adopted to classify the extracted features of the eating/drinking activities based on the space and time varying property of the features, by making use of the powerful modelling capability of HTM network on dynamic signals which is varying with both space and time. The proposed approach is tested through the real eating and drinking activities using the three dimensional accelerometers. Experimental results show that the EKF and HTM based two-stage approach can perform the activity detection successfully with very high accuracy.