Project Video: https://www.youtube.com/watch?v=0W8iLqHvZz8

Description: People spend a substantial part of life at rest in bed. 3D human pose and shape estimation for this activity would have numerous beneficial applications, such as patient care and assistive robotics. For example, a robot might be able to give a person a bed bath with detailed and accurate human pose information of the person in bed. However, line-of-sight perception is complicated by occlusion from bedding. Fabric pressure sensing mats are a promising alternative, which can be placed underneath a person in bed to sense the person. However, training data is challenging to collect at scale. We describe a physics-based method that simulates human bodies at rest in a bed with a pressure sensing mat, and present PressurePose, a synthetic dataset with 206K pressure images with 3D human poses and shapes. We also present PressureNet, a deep learning model that estimates human pose and shape from a pressure image and a measured height and weight.

Paper: https://arxiv.org/pdf/2004.01166.pdf