Wenwang Huang

Wenwang Huang 黄文旺

Ph.D. Student in Computer Science · Harbin Institute of Technology, Shenzhen
Supervised by Prof. Zhuotao Tian

I am a first-year Ph.D. student in Computer Science at Harbin Institute of Technology, Shenzhen. Before starting my Ph.D. study, I received my bachelor's degree in Computer Science and Technology from Shandong University. My research focuses on multimodal understanding and generation, reinforcement learning, and AI agents. My long-term goal is to genuinely advance AI intelligence and build AI assistants as capable and helpful as JARVIS. Feel free to reach out via email.

Email Google Scholar GitHub HuggingFace

News

Internships

Research Intern Tencent  ·  Research on reinforcement learning and agent-based paradigms for image and video generation.

Selected Publications

Multimodal Generation
DyRef
ECCV 2026Oral

Scaling Multi-Reference Image Generation with Dynamic Reward Optimization

Wenwang Huang*, Yusen Fu*, Junjie Wang, Mengfei Huang, Yulin Li, Gan Liu, Jing Cai, Yancheng He, Zhuotao Tian

European Conference on Computer Vision (ECCV), 2026

DyRef is a two-stage training framework for complex multi-reference image generation. It combines supervised fine-tuning with dynamic reward optimization to improve open-source models as the number and diversity of reference images increase. We release the DyRef model weights, the training data used in our work, and OmniRef-Bench, a dedicated benchmark for evaluating multi-reference image generation across diverse combinations of reference-image types and quantities.

Paper Code Model Weights Training Data Benchmark stars

Education

Harbin Institute of Technology, Shenzhen
Ph.D. Student in Computer Science
Advisor: Prof. Zhuotao Tian
Shandong University
B.Eng. in Computer Science and Technology
China

Honors & Awards

2024National Encouragement Scholarship for Undergraduate Students
2022National Encouragement Scholarship for Undergraduate Students

Academic Service

Academic service information will be updated here.