Yuanhan (John) Zhang

Hi! I'm Yuanhan Zhang; here is the standard Chinese pronunciation of my given name: Yuanhan. I am a third-year PhD student at MMLab@NTU, supervised by Prof. Ziwei Liu.

My research interests lie in computer vision and deep learning. I focus on adapting foundation models—from vision to multimodal—for real-world use: benchmarking model performance and adapting models via parameter-efficient tuning, in-context learning, and instruction tuning.

Email (yuanhan002@e.ntu.edu.sg)  /  Google Scholar  /  Twitter  /  GitHub

A holistic benchmark to assess advanced reasoning and understanding correctness/robustness between MLLMs and humans.

Fully open-sourced video LMM (code, model, and data) with competitive ability.

A family of LMMs consolidating insights into data, models, and visual representations.

Tackling multi-image, video, and 3D in large multimodal models.

Benchmarking 20 abilities of vision-language models.

An embodied VLM trained with RLEF, strong at embodied visual planning and programming.

Benchmarks funny, creative, and magic videos for challenging video understanding.

A vision-language model with in-context instruction tuning.

Retrieving prompts for visual in-context learning.

Learning without forgetting for vision-language models.

Searching prompt modules for parameter-efficient transfer learning.

3D point cloud pre-training with knowledge distillation from 2D images.

New benchmark for evaluating vision foundation models; supervised contrastive learning framework.

Large-scale face anti-spoofing dataset.

Last updated in Aug. 2025.

Homepage credits: Jon Barron.