Tracking People's Hands and Feet Using Mixed Network AND/OR Search

• Published in: IEEE transactions on pattern analysis and machine intelligence Vol. 35; no. 5; pp. 1248 - 1262
• Main Authors: Morariu, V. I., Harwood, D., Davis, L. S.
• Format: Journal Article
• Language: English
• Published: Los Alamitos, CA IEEE 01.05.2013; IEEE Computer Society
• Subjects: Applied sciences; Artificial intelligence; Basketball; Computer science; control theory; systems; Computer systems and distributed systems. User interface; Databases, Factual; Discriminant Analysis; Exact sciences and technology; Extremities; Foot - physiology; Graphical models; Hand - physiology; Humans; Image Processing, Computer-Assisted - methods; Information retrieval. Graph; Least-Squares Analysis; motion; Motor Activity - physiology; Pattern analysis; Pattern Recognition, Automated - methods; Pattern recognition. Digital image processing. Computational geometry; pictorial structures; Probabilistic logic; Search problems; Software; Theoretical computing; Tracking; Training; Occlusion; Lazy programming; High resolution; Tracking; Image processing; Video signal; Pedestrian; Activity; pictorial structures; Optimization; Multiple image; Robustness; Camera; Deterministic approach; Head; Target tracking; Image resolution; motion; Information retrieval; Foot; Multiple view; Implicit enumeration method; Optimal solution; Probabilistic net; Complex network; Occultation; Remote supervision
• DOI: 10.1109/TPAMI.2012.187
• ISSN: 0162-8828

We describe a framework that leverages mixed probabilistic and deterministic networks and their AND/OR search space to efficiently find and track the hands and feet of multiple interacting humans in 2D from a single camera view. Our framework detects and tracks multiple people's heads, hands, and feet through partial or full occlusion; requires few constraints (does not require multiple views, high image resolution, knowledge of performed activities, or large training sets); and makes use of constraints and AND/OR Branch-and-Bound with lazy evaluation and carefully computed bounds to efficiently solve the complex network that results from the consideration of interperson occlusion. Our main contributions are: 1) a multiperson part-based formulation that emphasizes extremities and allows for the globally optimal solution to be obtained in each frame, and 2) an efficient and exact optimization scheme that relies on AND/OR Branch-and-Bound, lazy factor evaluation, and factor cost sensitive bound computation. We demonstrate our approach on three datasets: the public single person HumanEva dataset, outdoor sequences where multiple people interact in a group meeting scenario, and outdoor one-on-one basketball videos. The first dataset demonstrates that our framework achieves state-of-the-art performance in the single person setting, while the last two demonstrate robustness in the presence of partial and full occlusion and fast nontrivial motion.