Hengshuang Zhao

International Hospital Kampala, University of Hong Kong

Papers

Total Citations

H-Index

About

Hengshuang Zhao is a researcher specializing in 3D computer vision, with a particular focus on unified object detection systems that bridge the gap between indoor and outdoor environments. His work addresses one of the field's most pressing challenges: developing models capable of generalizing across diverse scene types, a capability critical for real-world applications such as robot navigation and autonomous systems. Zhao's most notable contributions center on monocular 3D object detection, where he has pioneered unified frameworks that eliminate the need for separate, scene-specific models. His 2024 paper "UniMODE: Unified Monocular 3D Object Detection" has already garnered 19 citations, signaling rapid recognition within the computer vision community. Building on this foundation, his 2025 follow-up work further extends these principles to broader 3D detection paradigms through both algorithmic innovation and large-scale data unification strategies. What distinguishes Zhao's research is its practical ambition — tackling the real-world complexity of heterogeneous training data with significantly different geometric properties across scenes. His contributions represent meaningful progress toward robust, generalizable perception systems that could power the next generation of intelligent robots and autonomous vehicles.