Head Mounted

Head-mounted devices (HMDs) are revolutionizing human-computer interaction and data acquisition, enabling researchers to capture detailed egocentric data for various applications. Current research focuses on improving the accuracy and robustness of HMD-based motion capture and pose estimation, employing techniques like neural networks (including Transformers and diffusion models), visual-inertial odometry, and sophisticated sensor fusion methods to overcome challenges like sensor drift and occlusions. This work is driving advancements in fields ranging from virtual and augmented reality to robotics and healthcare, with significant implications for human-robot collaboration, surgical guidance, and accessibility of motion capture technology.