Mengqi Ji

Papers

Total Citations

H-Index

About

Mengqi Ji’s research lies at the intersection of computer vision, robotics, and human behavior understanding, with a focus on enabling intelligent systems to perceive and anticipate complex real-world interactions. Her most notable contribution is the **Group Interaction Field**, a novel framework for learning and explaining pedestrian anticipation in dense crowds. This work addresses a critical gap in autonomous navigation—how unmanned systems like service robots and self-driving cars can innately predict others’ actions, much like humans do. With 10 citations since its 2023 publication, this paper has quickly gained traction for its explainable approach to a traditionally black-box problem. Ji also advances **multimodal learning for surface material perception**, tackling the ill-posed challenge of disentangling material, lighting, and geometry in visual data. Her 2022 paper on this topic (7 citations) proposes a method that moves beyond appearance-based approaches, offering a more robust solution for robotic manipulation and scene understanding. By bridging high-level social reasoning with low-level physical perception, Ji’s work is shaping how autonomous systems navigate and interact with both people and objects—a dual contribution that positions her as a rising voice in embodied AI and human-robot interaction.