Max Robotics — Global Service Robot Directory + US Certification

About

Wei-Shi Zheng is a prominent computer vision and robotics researcher whose work spans two interconnected domains: human action recognition and robotic grasping. His most significant contributions center on advancing human-robot interaction (HRI) through the creation of large-scale RGB-D datasets capable of supporting arbitrary-view action recognition — a notoriously difficult challenge that single-view and multi-view approaches fail to address adequately. His database-building efforts, including the Varying-View RGB-D Action Dataset series published between 2018 and 2020, have collectively garnered over 130 citations and provided the research community with essential benchmarks for tackling real-world HRI scenarios. Zheng's robotics work demonstrates equal rigor, with contributions including TransGrasp, a transformer-based architecture for 7-DoF grasp detection that leverages multi-scale hierarchical point representations, and several innovative frameworks addressing cluttered-scene grasping, single-view depth reconstruction, and task-oriented manipulation. His more recent explorations into language-guided dexterous grasping and AI-assisted surgical robotics signal an expanding research vision that bridges embodied intelligence with clinical applications. With publications spanning foundational datasets to cutting-edge neural architectures, Zheng's work consistently tackles the gap between controlled laboratory settings and the complex, unpredictable demands of real-world robotic deployment.