CV
This is a description of the page. You can modify it in '_pages/cv.md'. You can also change or remove the top pdf download button.
Contact Information
| Name | Zhenyuan Zhang |
| Professional Title | Ph.D. Student |
| zzhangko@connect.ust.hk |
Professional Summary
Ph.D. student in Artificial Intelligence at HKUST, working on computer vision and deep learning.
Experience
-
2025 - Present Beijing, China
Research Assistant
GIGA
Supervised by Zheng Zhu. Working on motion generation and VLA.
- Contributed to Motion-R1, a text-to-motion generation framework combining Chain-of-Thought reasoning and reinforcement learning.
- Led MotionCoT Data Engine for automated motion-stage annotation via LLMs.
- Extended the GRPO method to motion synthesis.
-
2024 - 2025 Hong Kong
Research Assistant
The Hong Kong University of Science and Technology
Supervised by Prof. Wenhan Luo.
- Implemented real-time inference acceleration for talking head generation models using TensorRT and ONNX Runtime.
- Developed deep learning models for object detection (YOLO) and multi-object tracking (DeepSORT).
- Achieved real-time inference on NVIDIA Jetson Orin platforms using ROS.
- Enhanced multi-sensor fusion algorithms (LiDAR-camera fusion).
Education
-
2025 - Present Hong Kong
Ph.D.
The Hong Kong University of Science and Technology (HKUST)
Artificial Intelligence
- Supervisor: Prof. Yike Guo
-
2021 - 2025 Nanjing, Jiangsu, China
Ph.D. Student (Direct Entry)
Nanjing University of Science and Technology (NJUST)
Computer Science and Technology
- Supervisor: Prof. Jianfeng Lu
-
2017 - 2021 Nanjing, Jiangsu, China
B.Eng.
Nanjing University of Science and Technology (NJUST)
Software Engineering
- GPA: 3.48/4.00 (Rank: 4/114, Core Course GPA: 3.72/4.00)
- Supervisor: Prof. Jianfeng Lu
Skills
Programming Languages (): Python, C/C++, MATLAB
Frameworks & Tools (): PyTorch, ROS, TensorRT, ONNX, Docker, NumPy, SciPy, Pandas, Matplotlib
Languages
Chinese : Native speaker
English : IELTS 6.5
Interests
Computer Vision: Image Restoration, Face Restoration, Adversarial Robustness, Motion Generation, World Models