Qihui Li

Papers

Total Citations

H-Index

About

Qihui Li is a rising researcher in 3D computer vision and geometric deep learning, whose work focuses on advancing representations for point clouds and multi-view depth data. Their most notable contribution, "MD-Mamba: Feature extractor on 3D representation with multi-view depth" (2024), introduces a novel architecture that leverages state-space models to efficiently capture long-range dependencies in 3D data, achieving 4 citations and signaling early impact in this rapidly evolving field. Building on this, Li's 2025 paper "Enhancing point cloud feature representation via historical node state increments in graph neural networks" pioneers a technique that incorporates temporal node state updates into GNNs, improving feature learning for dynamic point cloud sequences. This work, already garnering 1 citation, demonstrates Li's ability to bridge graph neural networks with 3D perception challenges. By integrating multi-view depth cues with advanced sequence modeling, Li is carving out a niche at the intersection of efficient 3D feature extraction and graph-based learning—a promising direction for applications in autonomous navigation, robotics, and augmented reality.