Polarization Human Shape and Pose Dataset
This work addresses the need for specialized datasets to explore human shape estimation using polarization imaging, which is incremental as it builds on existing methods for surface normal reconstruction.
The authors tackled the problem of estimating detailed human body shapes by leveraging geometric cues from polarization images, resulting in the creation of the Polarization Human Shape and Pose Dataset (PHSPD).
Polarization images are known to be able to capture polarized reflected lights that preserve rich geometric cues of an object, which has motivated its recent applications in reconstructing detailed surface normal of the objects of interest. Meanwhile, inspired by the recent breakthroughs in human shape estimation from a single color image, we attempt to investigate the new question of whether the geometric cues from polarization camera could be leveraged in estimating detailed human body shapes. This has led to the curation of Polarization Human Shape and Pose Dataset (PHSPD), our home-grown polarization image dataset of various human shapes and poses.