Max Robotics — Global Service Robot Directory + US Certification

About

Shengping Zhang is a leading researcher in 3D computer vision and embodied AI, whose work bridges the critical gap between incomplete sensory data and robust real-world perception. His most impactful contribution is the **GRNet (Gridding Residual Network)** for dense point cloud completion, a seminal work that has garnered over 375 citations. By introducing a gridding-based architecture that preserves fine structural details—overcoming the limitations of traditional MLP-based methods like PCN—Zhang’s approach set a new standard for reconstructing complete 3D shapes from partial scans, with direct applications in robotics and autonomous systems. Beyond point clouds, Zhang has advanced **vision-and-language navigation (VLN)** with an object-and-action-aware model that enables robots to parse natural-language instructions and environmental cues simultaneously. His work also extends to **3D object reconstruction from stereo images**, improving generalization beyond training-set memorization, and to **weakly supervised object detection** in challenging underwater environments. More recently, he has pioneered **gait prediction for assistive robots**, developing the first end-to-end model that jointly predicts discrete locomotion modes and continuous joint kinematics across varied terrains. With a portfolio spanning 3D completion, embodied navigation, and human-robot interaction, Zhang’s research consistently pushes the boundaries of how machines perceive, reason, and act in the physical world.