"Let's get ready to bundle!": Crowd-level Human Keypoint Tracking

This work examines the suitability of a state-of-the-art human pose tracking method for application within surveillance scenarios and focuses on public places in urban areas that tend to suffer from crowdedness, such as city centers. Starting with a short introduction to motivate keypoint tracking in surveillance applications, this report will present details about the adapted method, which follows an LSTM-based approach. Afterwards, different changes that had to be incorporated in order to successfully apply the given method to our target setting will be presented. Finally, various experiments will show how the chosen method performs, based on experiments with simulated data.