|
Rongyao Fang 方荣耀方荣耀Researcher研究员
Qwen VL TeamQwen VL 团队
|
Biography个人简介
I am currently a Researcher at Alibaba Qwen VL Team, working on unified multimodal large models that seamlessly integrate visual understanding and generation. My research is driven by a passion for Artificial General Intelligence (AGI), with a focus on developing omni-modal foundation models that can perceive, reason, and create across vision, language, and beyond.
I obtained my Ph.D. from the Multimedia Laboratory (MMLab) of The Chinese University of Hong Kong (CUHK) in 2025, fortunate to be supervised by Prof. Hongsheng Li. I also worked closely with Prof. Xihui Liu.
Previously, I was a visiting scholar at MIT CSAIL, advised by Prof. Dina Katabi. I obtained my B.Eng. degree from Shanghai Jiao Tong University, where I was ranked 1st/157 and advised by Prof. Bingbing Ni.
我目前在阿里巴巴 Qwen VL 团队担任研究员，主要从事统一多模态大模型研究，推动视觉理解与视觉生成的一体化能力提升。我的研究以通用人工智能（AGI）为长期目标，关注全模态基础模型，让模型能够在视觉、语言等不同模态间实现统一的感知、推理与生成。
我于2025年在香港中文大学多媒体实验室（MMLab）获得博士学位，导师为李鸿升教授。博士期间，我也与刘希慧教授保持紧密合作。
此前，我曾在麻省理工学院计算机科学与人工智能实验室（MIT CSAIL）担任访问学者，导师为 Prof. Dina Katabi。本科毕业于上海交通大学信息工程专业，排名第1/157名，导师为倪冰冰教授。
News最新动态
- [Jan. 2026] Two papers accepted to ICLR 2026.
- [Sep. 2025] One paper accepted to NeurIPS 2025.
- [Jul. 2025] One paper accepted to ICCV 2025.
- [Feb. 2025] One paper accepted to CVPR 2025.
- [2026年1月] 两篇论文被 ICLR 2026 接收。
- [2025年9月] 一篇论文被 NeurIPS 2025 接收。
- [2025年7月] 一篇论文被 ICCV 2025 接收。
- [2025年2月] 一篇论文被 CVPR 2025 接收。
Education教育背景
- [2021 - 2025] Ph.D. at MMLab, The Chinese University of Hong Kong.
- [2016 - 2020] B.Eng. in Information Engineering, Shanghai Jiao Tong University (Ranking: 1st/157).
- [2019 - 2020] Visiting Scholar at CSAIL, Massachusetts Institute of Technology.
- [2021 - 2025] 香港中文大学多媒体实验室（MMLab）博士。
- [2016 - 2020] 上海交通大学信息工程专业工学学士（排名：第1/157名）。
- [2019 - 2020] 麻省理工学院计算机科学与人工智能实验室（CSAIL）访问学者。
Publications学术论文
(* indicates equal contribution)(* 表示同等贡献)
-
MathCanvas: Intrinsic Visual Chain-of-Thought for Multimodal Mathematical Reasoning
-
CodePlot-CoT: Mathematical Visual Reasoning by Thinking with Code-Driven ImagesarXiv preprint, 2025 [Paper]
-
FLUX-Reason-6M & PRISM-Bench: A Million-Scale Text-to-Image Reasoning Dataset and Comprehensive BenchmarkInternational Conference on Learning Representations (ICLR), 2026 [Paper]
-
T2I-ReasonBench: Benchmarking Reasoning-Informed Text-to-Image GenerationarXiv preprint, 2025 [Paper]
-
GoT-R1: Unleashing Reasoning Capability of MLLM for Visual Generation with Reinforcement LearningInternational Conference on Learning Representations (ICLR), 2026 [Paper]
-
GoT: Unleashing Reasoning Capability of Multimodal Large Language Model for Visual Generation and EditingConference on Neural Information Processing Systems (NeurIPS), 2025 [Paper]
-
CrossLMM: Decoupling Long Video Sequences from LMMs via Dual Cross-Attention MechanismsarXiv preprint, 2025 [Paper]
-
SOLVE: Synergy of Language-Vision and End-to-End Networks for Autonomous DrivingConference on Computer Vision and Pattern Recognition (CVPR), 2025 [Paper]
-
StreamChat: Chatting with Streaming VideoarXiv preprint, 2024 [Paper]
-
Puma: Empowering Unified MLLM with Multi-Granular Visual GenerationInternational Conference on Computer Vision (ICCV), 2025 [Paper]
-
Mimic Before Reconstruct: Enhancing Masked Autoencoders with Feature MimickingInternational Journal of Computer Vision (IJCV), 2024 [Paper]
-
FeatAug-DETR: Enriching One-to-Many Matching for DETRs with Feature AugmentationTransactions on Pattern Analysis and Machine Intelligence (TPAMI), 2024 [Paper]
-
FouriScale: A Frequency Perspective on Training-Free High-Resolution Image SynthesisEuropean Conference on Computer Vision (ECCV), 2024 [Paper]
-
Clip-Adapter: Better Vision-Language Models with Feature AdaptersInternational Journal of Computer Vision (IJCV), 2024 [Paper]
-
InstructSeq: Unifying Vision Tasks with Instruction-Conditioned Multi-Modal Sequence GenerationarXiv preprint, 2023 [Paper]
-
Point-M2AE: Multi-Scale Masked Autoencoders for Hierarchical Point Cloud Pre-TrainingConference on Neural Information Processing Systems (NeurIPS), 2022 [Paper]
-
RBGNet: Ray-Based Grouping for 3D Object DetectionConference on Computer Vision and Pattern Recognition (CVPR), 2022 [Paper]
-
Tip-Adapter: Training-Free CLIP-Adapter for Better Vision-Language ModelingEuropean Conference on Computer Vision (ECCV), 2022 [Paper]
-
Learning Longterm Representations for Person Re-Identification Using Radio SignalsConference on Computer Vision and Pattern Recognition (CVPR), 2020 [Paper]
-
Probabilistic Radiomics: Ambiguous Diagnosis with Controllable Shape AnalysisInternational Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI), 2019 [Paper]
-
Adversarial Attack and Defense on Point SetsarXiv preprint, 2019 [Paper]
Experience工作经历
- [Sep. 2025 - Present] Researcher, Alibaba Qwen VL Team.
- [Feb. 2024 - Aug. 2025] Research Intern, SenseTime.
- [Jun. 2022 - Apr. 2023] Research Intern, Shanghai AI Laboratory.
- [2025年9月 - 至今] 阿里巴巴 Qwen VL 团队，研究员。
- [2024年2月 - 2025年8月] 商汤科技，研究实习生。
- [2022年6月 - 2023年4月] 上海人工智能实验室，研究实习生。
Selected Awards所获荣誉
- [2021] Hong Kong PhD Fellowship.
- [2020] Outstanding Graduates of Shanghai (Top 1%).
- [2017 & 2018] National Scholarship (Top 1%).
- [2017 & 2018] Zhiyuan College Honors Scholarship (Top 5%).
- [2021] 香港政府博士奖学金（Hong Kong PhD Fellowship）。
- [2020] 上海市优秀毕业生（前 1%）。
- [2017 & 2018] 国家奖学金（前 1%）。
- [2017 & 2018] 致远学院荣誉奖学金（前 5%）。