Photo from Scientific Research

Visualization of different poses with the same character. Top: poses rendered from an external camera viewpoint. White represents occlusion, which is body parts that would not be visible from the egocentric perspective. Bottom: poses rendered from the egocentric camera viewpoint. Color gradient indicates the density of image pixels for each area of the body: green is higher pixel density, whereas red is lower density. This figure illustrates the challenges faced in egocentric human pose estimation: severe self-occlusions, extreme perspective effects and lower pixel density for the lower body.

DOI