Parallelization strategies for markerless human motion capture

• Published in: Journal of real-time image processing Vol. 14; no. 2; pp. 453 - 467
• Main Authors: Cano, Alberto, Yeguas-Bolivar, Enrique, Muñoz-Salinas, Rafael, Medina-Carnicer, Rafael, Ventura, Sebastián
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.02.2018; Springer Nature B.V
• Subjects: Algorithms; Camcorders; Central processing units; Computer Graphics; Computer Science; Computing time; CPUs; Genetic algorithms; Human motion; Image Processing and Computer Vision; Motion capture; Multimedia Information Systems; Optimization; Original Research Paper; Pattern Recognition; Performance evaluation; Signal,Image and Speech Processing; Markerless motion capture (MMOCAP); GPU; Tracking
• ISSN: 1861-8200; 1861-8219
• DOI: 10.1007/s11554-014-0467-1

Markerless motion capture (MMOCAP) is the problem of determining the pose of a person from images captured by one or several cameras simultaneously without using markers on the subject. Evaluation of the solutions is frequently the most time-consuming task, making most of the proposed methods inapplicable in real-time scenarios. This paper presents an efficient approach to parallelize the evaluation of the solutions in CPUs and GPUs. Our proposal is experimentally compared on six sequences of the HumanEva-I dataset using the CMAES algorithm. Multiple algorithm’s configurations were tested to analyze the best trade-off with regard to the accuracy and computing time. The proposed methods obtain speedups of 8 × in multi-core CPUs, 30 × in a single GPU and up to 110 × using 4 GPUs.