Hao Xing

Hao Xing is a scientific researcher and post-doctoral fellow at the Institute for Cognitive Systems (ICS), Technical University of Munich (TUM), working with Prof. Gordon Cheng since May 2024. Previously, he served as a research assistant at the Munich Institute of Robotics and Machine Intelligence (MIRMI) and completed his PhD under Prof. Darius Burschka from 2019.

His educational background includes:

• Master of Science in Mechanical Engineering from Technical University of Munich
• Bachelor of Engineering in Mechanical Engineering from Hefei University of Technology, China

Xing's research focuses on Robot Vision, Human Action Recognition, and Graph Convolutional Networks, with significant contributions to Human-Object Interaction Recognition, Scene Understanding, and Visual Depth Estimation. His work leverages machine learning to develop robust algorithms for human activity analysis and robotic perception, emphasizing spatio-temporal modeling and uncertainty handling in real-world scenarios.

Analysis of his 13 recent publications (2019-2025) reveals a dominant focus on graph-based approaches for action recognition and segmentation, with increasing emphasis on open-world applications, uncertainty modeling, and healthcare integration. His research spans computer vision, robotics, and medical applications, demonstrating strong interdisciplinary impact through collaborations in surgical robotics and patient monitoring systems.

Xing actively mentors students through master's thesis projects in Stereo Matching, Human Activity Segmentation, and Monocular Depth Estimation, requiring expertise in deep learning frameworks and computer vision algorithms. While no major grants are explicitly mentioned, his thesis topics indicate active research funding in geometric vision and human activity analysis.

He operates within the Institute for Cognitive Systems (ICS) at TUM, a key unit in MIRMI's robotics ecosystem focused on advancing human-robot interaction through vision-based scene understanding and motion generation capabilities.