Xiaozheng Liu

Papers

Total Citations

H-Index

About

Xiaozheng Liu is a robotics researcher whose work centers on enabling robots to perceive and manipulate objects in unstructured environments, with a particular focus on semantic segmentation and grasp pose detection. His major contributions lie in advancing few-shot learning for robotic grasping, where he pioneered methods to segment unseen objects with minimal training data—a critical challenge for real-world deployment. His 2022 paper on "Unseen Object Few-Shot Semantic Segmentation for Robotic Grasping" (15 citations) introduced a novel framework that allows robots to adapt to novel environments without exhaustive retraining. Building on this, Liu developed multi-modal fusion techniques that integrate RGB and depth data, as demonstrated in his 2023 work on "Joint Segmentation and Grasp Pose Detection with Multi-Modal Feature Fusion Network" (7 citations), which outperformed single-modality approaches in cluttered scenes. His TRF-Net (5 citations) further advanced transformer-based architectures for desktop object segmentation. Liu’s research directly addresses the gap between laboratory-controlled settings and real-world robotic applications, making his work highly relevant for researchers in manipulation, computer vision, and embodied AI.