I am a second-year CS Ph.D. student at the University of Washington, advised by Prof. Ranjay Krishna. I am also a student researcher at the Allen Institute for AI. Before UW, I was an undergraduate at the University of Michigan and Shanghai Jiao Tong University, where I was fortunate to also work with Prof. Andrew Owens and Dr. Adam Harley. My primary research interest lies in computer vision and multimodal learning.
I am generally interested in various problems in world modeling and vision-language models. Here are my current research interests:
1) Interactive video models. I am increasingly interested in interaction models that perceive video streams in real time, anticipate what happens next, and engage with people as capable AI assistants [1, 2]. I care about both the modeling challenge in a real-time, full-duplex setting and the challenge of building faithful environments to train and evaluate such interactive agents. I am also interested in making world models compact yet visually interpretable so they fit in this streaming & interactive setting (see MolmoMotion).
2) Multimodal pretraining. I have a long-standing interest in visual pretraining and how it fits into modern multimodal pretraining. My work in this area includes novel tokenization (TrajViT, TrajTok) and multimodal representation learning (EXIF as Language, iterated learning for VLMs). I currently co-lead an open-source vision encoder project at AI2.
Meta AI Research (FAIR)Incoming, Starting from Sept. 2026
Visiting Researcher
Salesforce AI Research, June 2026 ~ Sept. 2026
Research Scientist Intern
Host:
Juan Carlos Niebles and
Silvio Savarese
Allen Institute for AI, Jan. 2025 ~ June 2026
Student Researcher (also Research Intern, Summer 2025)
MolmoMotion: Forecasting Point Trajectories in 3D with Language Instruction
Jianing Zhang*, Chenhao Zheng*, Yajun Yang, Rustin Soraki, Winson Han, Chun-Liang Li, Jason Ren, Max Argus, Jieyu Zhang, Ranjay Krishna
In submission
paper
·
Blog at AI2
·
project page
TrajTok: Learning Trajectory Tokens Enhances Video Understanding
Chenhao Zheng, Jieyu Zhang, Jianing Zhang, Weikai Huang, Ashutosh Kumar, Quan Kong, Oncel Tuzel, Chun-Liang Li, Ranjay Krishna
CVPR 2026
paper
One Trajectory, One Token: Grounded Video Tokenization via Panoptic Sub-object Trajectory
Chenhao Zheng, Jieyu Zhang, Reza Salehi, Ziqi Gao, Vishnu Iyengar, Norimasa Kobori, Quan Kong, Ranjay Krishna
ICCV 2025 (Highlight)
project page
·
paper
Synthetic Visual Genome
Jae Sung Park, Zixian Ma, Linjie Li, Chenhao Zheng, Cheng-Yu Hsieh, Ximing Lu, Khyathi Chandu, Quan Kong, Norimasa Kobori, Ali Farhadi, Yejin Choi, Ranjay Krishna
CVPR 2025
project page
·
paper
Acoustic Volume Rendering for Neural Impulse Response Field
Zitong Lan, Chenhao Zheng, Zhiwei Zheng, Mingmin Zhao
NeurIPS 2024 (Spotlight)
project page
·
paper
Iterated Learning Improves Compositionality in Large Vision-Language Models
Chenhao Zheng, Jieyu Zhang, Aniruddha Kembhavi, Ranjay Krishna
CVPR 2024
project page
·
paper
·
video
EXIF as Language: Learning Cross-Modal Associations Between Images and Camera Metadata
Chenhao Zheng, Ayush Shrivastav, Andrew Owens
CVPR 2023 (Highlight)
project page
·
paper
·
video
A VR capstone project from my undergrad at Michigan that I really love.
Urban Rush: Making Fitness Fun with VR
with Rahmy Salmon,
Kalpit Haresh Sutariya,
Akash Nallani, and
Ashish Patel.
An immersive VR experience encouraging exercise via a narrative-based exploration in a procedurally generated city.
project page