English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

A classical approach to abnormal activity detection is to learn arepresentation for normal activities from the training data and then use thislearned representation to detect abnormal activities while testing. Typically,the methods based on this approach operate at a fixed timescale - either asingle time-instant (eg. frame-based) or a constant time duration (eg.video-clip based). But human abnormal activities can take place at differenttimescales. For example, jumping is a short term anomaly and loitering is along term anomaly in a surveillance scenario. A single and pre-definedtimescale is not enough to capture the wide range of anomalies occurring withdifferent time duration. In this paper, we propose a multi-timescale model tocapture the temporal dynamics at different timescales. In particular, theproposed model makes future and past predictions at different timescales for agiven input pose trajectory. The model is multi-layered where intermediatelayers are responsible to generate predictions corresponding to differenttimescales. These predictions are combined to detect abnormal activities. Inaddition, we also introduce an abnormal activity data-set for research use thatcontains 4,83,566 annotated frames. Data-set will be made available athttps://rodrigues-royston.github.io/Multi-timescale_Trajectory_Prediction/ Ourexperiments show that the proposed model can capture the anomalies of differenttime duration and outperforms existing methods.

Multi-timescale Trajectory Prediction for Abnormal Human Activity Detection