Indoor Body Perception

Indoor body perception research aims to accurately understand and represent the human body within indoor environments using various sensing modalities, primarily focusing on improving the accuracy and efficiency of 3D body reconstruction and pose estimation. Current efforts leverage advanced techniques like physics-informed generative neural networks to model radio-frequency propagation and convolutional neural networks for depth map completion from sparse sensor data, often incorporating perceptual grouping for improved feature extraction. These advancements are driving progress in applications such as robotic navigation, elderly care, and building energy management by enabling more robust and efficient human-computer interaction within indoor spaces.