Dongfang Liu

Rochester Institute of Technology, Purdue University West Lafayette

Papers

Total Citations

H-Index

About

Dongfang Liu is a researcher advancing the frontier of embodied AI and robotic perception, with a focus on enabling machines to interact intelligently with complex, real-world environments. His key research areas span deep robotic grasping, multimodal navigation, and visual localization for field robots. Liu’s most cited work, “Deep Robotic Grasping Prediction with Hierarchical RGB-D Fusion” (2022, 43 citations), introduces a novel hierarchical fusion architecture that integrates RGB and depth data to significantly improve grasping accuracy—a critical capability for autonomous manipulation. In his innovative “Multimodal Aggregation Approach for Memory Vision-Voice Indoor Navigation with Meta-Learning” (2020, 20 citations), Liu tackles the challenge of agent-environment interaction by combining visual observation with voice commands, enhanced by meta-learning for adaptive navigation. He also contributes to robust visual localization with his “Triangulation-Based Visual Localization for Field Robots” (2022, 16 citations), which improves GPS-denied positioning for outdoor robots. Collectively, Liu’s work demonstrates a strong impact in robotics and computer vision, bridging perception and action to create more capable, autonomous systems.