Hi, my name is

Zherui Qiu.

I Empower Robots to Perceive, Imagine, and Execute

A passionate and innovative AI researcher. I design agents that perceive the world, imagine outcomes, and execute them with reliable control. I also harness cutting-edge techniques to craft virtual worlds that are visually stunning and immersive.

"To boldly go where no one has gone before." — Star Trek

About Me

I am a Ph.D. student in Data Science at the University of Science and Technology of China advised by Prof. Juyong Zhang. My research includes publications on embodied intelligence, reinforcement learning, and Neural Radiance Fields (NeRF). Currently, I am focused on applying these techniques to embodied intelligence, aiming to enhance spatial understanding and interaction within complex environments.

Beyond my academic pursuits, I actively participate in collaborative projects and workshops, seeking to stay at the forefront of technological advancements and foster interdisciplinary research.

Here are a few technologies I've been working with recently:

VLA Model
World Model
Diffusion Policy
CUDA C++
NeRF/3D-GS
ComfyUI

Experience

Shanghai AI Lab
OpenDriveLab

Research Intern (Full-Time) - Shanghai AI Lab

June 2025 - Present

I am currently working as a research intern at Shanghai AI Lab in Shanghai, China.
On the research side, as a core contributor to InternVLA-A1, I have learned how to design robust VLA architectures, build scalable training pipelines, and run principled evaluations, while sharpening my intuition for data curation and ablation analysis.
On the engineering side, building ROSGUI-AGX, a modern GUI for orchestrating ROS commands and teleoperation-based data collection with live process monitoring, has taught me how to turn collection protocols into operator-friendly workflows, provide real-time feedback and logging, and keep long-running robotics processes stable and reproducible.

Research Intern (Full-Time) - OpenDriveLab

Nov 2024 - May 2025

I worked as a research intern at OpenDriveLab in Shanghai, China. I researched embodied intelligence by integrating advanced algorithms with robotic systems and refining manipulation strategies through hands-on experimentation. I also assisted in comprehensive data collection, analysis, and documentation to ensure accurate reporting and clear presentation of experimental findings.

Education

2020.09 - Present

Ph.D. of Data Science

School of Artificial Intelligence and Data Science, University of Science and Technology of China

GPA: 3.71 out of 4.3

I have authored and co-authored 6 top‑tier papers in embodied AI and computer graphics.

🏅 InternVLA-A1: Unifying Understanding, Generation and Action for Robotic Manipulation. Preprint 2026.
🏅 Deformable NeRF using Recursively Subdivided Tetrahedra. ACM Multimedia 2024.
🤝 InternData-A1: Pioneering High-Fidelity Synthetic Data for Pre-training Generalist Policy. Preprint 2025.
🤝 F1: A Vision-Language-Action Model Bridging Understanding and Generation to Actions. Preprint 2025.
🤝 Promoting Stochasticity for Expressive Policies via a Simple and Efficient Regularization Method. NeurIPS 2020.
🤝 PaletteGaussian: 3D Photorealistic Color Editing with Gaussian Splatting. ISMAR 2024.