About

Jiazhao Zhang is a robotics and embodied AI researcher whose work spans robotic perception, manipulation, and navigation — areas at the intersection of computer vision, 3D scene understanding, and autonomous systems. He is perhaps best known for **GraspNeRF** (2023, 91 citations), a pioneering framework that leverages generalizable Neural Radiance Fields to enable 6-DoF grasp detection for transparent and specular objects — a notoriously difficult challenge where conventional depth cameras fail. This contribution marked the first multiview RGB-based solution to this problem, earning significant recognition from the robotics community. His earlier work on active scene understanding through online semantic reconstruction (2019, 38 citations) demonstrated his long-standing interest in intelligent robot exploration and semantic 3D mapping. More recently, Zhang has expanded into mobile manipulation with GAMMA (2024, 20 citations), addressing real-time grasping pose estimation during robot approach, and into vision-and-language navigation through the NaVid series, which incorporates video-based reasoning and 4D spatial intelligence. His 2025 platform RoboVerse further reflects a growing commitment to scalable, generalizable robot learning infrastructure. Across these contributions, Zhang has established himself as a versatile and impactful voice in modern embodied AI research.

Research Focus

Key Achievements

4
H-Index
8
Papers
162
Total Citations
20
Avg Citations/Paper
🏆 Most Cited Paper
GraspNeRF: Multiview-based 6-DoF Grasp Detection for Transparent and Specular Objects Using Generalizable NeRF
91 citations · 2023
📈 Most Prolific Year: 2019 (2 Papers)
🤝 Key Collaborators: 58
🏛 Institutions: Beijing Academy of Artificial Intelligence, National University of Defense Technology, Peking University, King University

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 15 days ago