Wenlong Dong

Southern University of Science and Technology

Papers

Total Citations

H-Index

About

Wenlong Dong is a leading researcher in robotic manipulation, with a primary focus on task-oriented grasping (TOG)—the critical ability for robots to grasp objects in a way that is compatible with a downstream task, such as using a tool. His work bridges the gap between semantic understanding and geometric reasoning, drawing inspiration from how the human brain activates distinct regions for these functions. Dong’s major contributions include the development of FoundationGrasp, a generalizable framework that leverages foundation models to synthesize task-compatible grasps without costly manual annotations. This work, published in 2025, has already garnered 22 citations, reflecting its immediate impact on the field. He also introduced RTAGrasp, a novel method that learns TOG from human videos through retrieval, transfer, and alignment, enabling robots to extract both grasp positions and directions from natural demonstrations. Dong’s research is notable for its practical approach to reducing annotation costs while improving generalization, making TOG more accessible for real-world robotic applications. His work is foundational for advancing autonomous tool manipulation and human-robot collaboration.