VIBE (video inference for body pose and shape)
VIBE (video inference for body pose and shape) is a computational framework designed to estimate 3D human body pose and shape from monocular video input. It employs deep learning techniques to produce temporally coherent and accurate reconstructions of human motion and body geometry in real-time.