Max Robotics — Global Service Robot Directory + US Certification

About

Xun Xu is a researcher whose work sits at the dynamic intersection of computer vision, robotics, and machine learning, with a particular focus on enabling intelligent machines to perceive and interact with the physical world. His most recognized contribution, "3D AffordanceNet: A Benchmark for Visual Object Affordance Understanding" (2021), has accumulated over 115 citations and represents a landmark advancement in visual affordance research — extending affordance understanding from 2D and 2.5D image domains into fully three-dimensional environments, providing the community with a robust benchmark for categorizing, segmenting, and reasoning about object interactions. This work has become an essential reference for vision-guided robotics research. Beyond affordance understanding, Xu has made meaningful contributions to the adaptation of large-scale foundation models, exploring how segmentation models like Segment-Anything (SAM) can be improved under distribution shifts via weakly supervised techniques. His more applied work addresses human-robot collaboration safety, proposing deep reinforcement learning-based trajectory planning for dynamic obstacle avoidance in manufacturing settings. Together, these contributions reflect a researcher committed to bridging fundamental perception research with real-world robotic applications, consistently pushing the boundaries of how machines understand and safely navigate human environments.