ADMA 2019 Competition

The ADMA 2019 Competition will be “Robust Automated Event Detection for Mult-Camera Streaming Video Environment”. This competition intends to develop robust automated activity detection for a multi-camera streaming video environment. As an essential aspect of this competition, activities will be enriched by person and object detection, as well as recognition at multiple levels of granularity.

This competition will focus on three major thrusts: 1) Detection of primitive activities occurring in ground-based video collection; Examples include: (1) Person getting into a vehicle, (2) Person getting out of a vehicle, (3) Person carrying object.

2) Detection of complex activities, including pre-specified or newly defined activities; Examples include: (1) Person being picked up by vehicle, (2) Person abandoning object, (3) Two people exchaning an object, (4) Person carrying a firearm.

3) Person and object detection and recognition across multiple overlapping and nonoverlapping camera viewpoints.

This competition will produce a common framework and software prototype for activity detection, person/object detection and recognition across a multi-camera network. The impact will be the development of tools for forensic analysis, as well as real-time alerting for userdefined threat scenarios.

— How to Participate

To take part in this competition, you need to send an email to adma2019contest@gmail.com to register and then acknowledge that you have read and accepted the terms and conditions.

The ADMA 2019 competition team will send you information on how to download data. Then you will email your JSON formatted to the competition team. The team will evaluate your result, and update on the leaderboard.

— Judging Criteria and Metrics

The main scoring metrics was based on activity detection using the weighted probability of missed detection at 0.15 false alarms per minute averaged over activity types. You can find the scoring software here https://github.com/usnistgov/ActEV_Scorer .

— Task coordinator

ADMA 2019 Competition team (adma2019contest@gmail.com)