Detection and Tracking of Occluded People

• Published in: International journal of computer vision Vol. 110; no. 1; pp. 58 - 69
• Main Authors: Tang, Siyu, Andriluka, Mykhaylo, Schiele, Bernt
• Format: Journal Article
• Language: English
• Published: Boston Springer US 01.10.2014; Springer; Springer Nature B.V
• Subjects: Artificial Intelligence; Benchmarks; Computer Imaging; Computer Science; Computer vision; Datasets; Detectors; Hypotheses; Image Processing and Computer Vision; Occlusion; Pattern Recognition; Pattern Recognition and Graphics; Sensors; Single persons; State of the art; Streets; Studies; Tracking; Vision; Vision systems; Multiple people tracking; Occlusion handling; Tracking; Pedestrian detection
• ISSN: 0920-5691; 1573-1405
• DOI: 10.1007/s11263-013-0664-6

We consider the problem of detection and tracking of multiple people in crowded street scenes. State-of-the-art methods perform well in scenes with relatively few people, but are severely challenged by scenes with many subjects that partially occlude each other. This limitation is due to the fact that current people detectors fail when persons are strongly occluded. We observe that typical occlusions are due to overlaps between people and propose a people detector tailored to various occlusion levels. Instead of treating partial occlusions as distractions, we leverage the fact that person/person occlusions result in very characteristic appearance patterns that can help to improve detection results. We demonstrate the performance of our occlusion-aware person detector on a new dataset of people with controlled but severe levels of occlusion and on two challenging publicly available benchmarks outperforming single person detectors in each case.