About Me

Hi there! I am a fourth-year Ph.D. student in Computer Science at Vanderbilt University, advised by Prof. Daniel Moyer. Prior to this, I received my B.Sc. (Hons.) in Computer Science from Hong Kong Baptist University, where I was granted the Scholastic Award as one of the top 2% of graduates.

My research interests primarily lie in Generative Models, Model Inversion, Membership Inference, and Monocular Depth Estimation. I am broadly interested in adversarial attacks and defenses. My current research focuses on inverting training datasets from pre-trained generative models, specifically diffusion-based and flow-matching architectures. I am also working as a Research Assistant for the Advanced Research Projects Agency for Health (ARPA-H), focusing on robotic vision development—specifically Monocular Depth Estimation (MDE) for endoscopic imaging in automated surgical programs.

I am actively seeking internship opportunities for both Summer and Fall 2026. Please feel free to drop me an email at mingxing.rao@vanderbilt.edu if our research interests align, or if you would like to collaborate!

🔥 Recent News

• [2026] Our paper “Latent Diffusion Inversion Requires Understanding the Latent Space” has been accepted to CVPR 2026!
• [2026] Our paper “Generalization and Memorization in Rectified Flow” is currently under review at ECCV 2026.
• [2024] Our paper “Zero-shot Prompt-based Video Encoder for Surgical Gesture Recognition” was accepted at IPCAI 2024 and published in IJCARS.

🌟 Beyond Research (Side Quests & Hobbies)

• AI Agents: Raising a squad of four OpenClaw to automate my daily routine. Still exploring more possibilities!
• My “Secondary Research”: I am a huge car enthusiast. Whether it’s geeking out over vehicle dynamics, following Formula 1, or just hitting the road to decompress, fun driving is my ultimate stress relief. Always down to chat about anything automotive!