I am a master’s student in Informatics at the Technical University of Munich, working on computer vision, graphics and robotics.
I am now a research assistant at the TUM Computer Vision Group, advised by Maolin Gao and supervised by Prof. Daniel Cremers. I also collaborate closely with Kaixin Bai and Lei Zhang at the Agile Robots.
I graduate from Technical University of Berlin with a bachelor’s degree in Computer Engineering, advised by Dr. Linh Kästner and supervised by Prof. Jens Lambrecht.
My current research focuses on computer vision, shape analysis, and robotic simulation, aiming to bridge the gap between simulation and reality through geometric and learning-based methods.
If you are seeking any form of academic cooperation, feel free to drop me an email at huajian.zeng@tum.de.
🔥 News
- 2025.01: 🎉Our paper on Arena-Rosnav 4.0 platform got accepted in ICRA 2025 in Atlanta, USA.
- 2024.11: 🎉Our paper on embedding learning for point clouds got accepted in 3DV 2025 in Singapore.
- 2024.04: 🎉Our paper on Arena-Rosnav 3.0 platform got accepted in RSS 2024 in Delft, Netherlands.
- 2023.07: I join Agile Robots
as a compter vision intern in Munich. - 2023.06: 🎉Our paper on Arena-Rosnav 2.0 platform got accepted in IROS 2023 in Detroit, USA.
- 2023.04: 🎉Our paper on 2D DRL-based robot navigation simulator got accepted in Ubiquitous Robots 2023 in Honolulu, USA.
- 2022.10: Move to Munich and started my master’s study at TUM.
📝 Publications
CoE: Deep Coupled Embedding for Non-Rigid Point Cloud Correspondences
Huajian Zeng, Maolin Gao, Daniel Cremers
- a novel deep coupled embedding learning method for dense non-rigid point cloud matching.
Volodymyr Shcherbyna, Linh Kästner, Diego Diaz, Huu Giang Nguyen, Maximilian Ho-Kyoung Schreff, Tim Seeger, Jonas Kreutz, Ahmed Martban, Huajian Zeng, Harold Soh
- A platform, integrating generative-model-based dynamic world generation, a semantically linked 3D model database, and extended human behavior simulation.
ClearDepth: Enhanced Stereo Perception of Transparent Objects for Robotic Manipulation
Kaixin Bai, Huajian Zeng, Lei Zhang, Yiwen Liu, Hongli Xu, Zhaopeng Chen, Jianwei Zhang
- Simulation data generation for photorealistic datasets of transparent objects based on Sim2Real.
- Stereo imaging algorithm for scene reconstruction of transparent objects using a novel cascaded ViT model
Arena 3.0: Advancing Social Navigation in Collaborative and Highly Dynamic Environments
Linh Kästner, Reyk Carstens, Huajian Zeng, Jacek Kmiecik, Tuan Anh Le, Teham Bhuiyan, Boris Meinardus, Jens Lambrecht
- A modular platform focused on human behavior modeling realism by integrating diverse social force models, task generation tools, and cross-simulator functionality abstraction.
Linh Kästner, Reyk Carstens, Huajian Zeng, Jacek Kmiecik, Tuan Anh Le, Teham Bhuiyan, Boris Meinardus, Jens Lambrecht
- A platform for developing and benchmarking robot navigation in highly dynamic environments.
Efficient 2D Simulators for Deep-Reinforcement-Learning-based Training of Navigation Approaches
Huajian Zeng, Linh Kästner, Jens Lambrecht
- Efficient 2D simulator for deep reinforcement learning (DRL)-based autonomous agent.
🎖 Honors and Awards
- 2024.10 Deutschestipendium 2024/2025
📖 Educations
- 2022.09 - 2025.03 (expected), Master, Informatics, Technical University of Munich, Munich.
- 2019.10 - 2022.09, Undergraduate, Computer Engineering, Technical University of Berlin, Berlin.
💻 Internships
- 2023.07 - 2024.05, Agile Robots, Munich.
🤝 Volunteer Works
- 2024.06 - 2024.07, UEFA Euro 2024, Munich.