Parallel deep learning ensembles for human pose estimation
ASME Dynamic Systems and Control Conference, 2018
Recommended citation: Ren, Hailin, Anil Kumar, Xinran Wang, and Pinhas Ben-Tzvi. “Parallel deep learning ensembles for human pose estimation”. In Dynamic Systems and Control Conference, vol. 51890, p. V001T07A005. American Society of Mechanical Engineers, 2018. https://asmedigitalcollection.asme.org/DSCC/proceedings/DSCC2018/51890/V001T07A005/455561
This paper presents an efficient method to detect human pose with monocular color imagery using a parallel architecture based on deep neural network. The network presented in this approach consists of two sequentially connected stages of 13 parallel CNN ensembles, where each ensemble is trained to detect one specific kind of linkage of the human skeleton structure. After detecting all skeleton linkages, a voting score-based post-processing algorithm assembles the individual linkages to form a complete human structure.