PeopleofSuperintelligence

Researcher

Lu Yuan

Lu Yuan is a computer-vision researcher whose papers cover vision transformers, grounded language-image pre-training and the Florence family of vision foundation models. He earned an MS at Tsinghua University in 2005 and a PhD at the Hong Kong University of Science and Technology in 2009, and he worked in the Visual Computing Group at Microsoft Research Asia from 2009 as a senior research manager. He coauthored CvT in March 2021 and was first author of Florence in November 2021, then coauthored GLIP in 2022, the Phi-3 technical report in April 2024 and Florence-2 in June 2024; Microsoft Research listed him as a Principal Research Manager on its Project Florence-VL team. As of October 2026, he is a Research Scientist Manager in the GenAI group at Meta.

Now at
Meta
Badges

Timeline

  1. 2005–2009

    Tsinghua and HKUST degrees

    He earns an MS from the Department of Electronic Engineering at Tsinghua University in 2005 and a PhD from the Department of Computer Science and Engineering at the Hong Kong University of Science and Technology in 2009.

  2. 2009

    Visual Computing at Microsoft Research Asia

    From 2009, he works in the Visual Computing Group at Microsoft Research Asia and holds the title of senior research manager there.

  3. March 2021

    The CvT vision transformer

    In March 2021, he coauthors CvT with Haiping Wu, Bin Xiao, Noel Codella, Mengchen Liu, Xiyang Dai and Lei Zhang. The Convolutional vision Transformer puts convolutions into the Vision Transformer and reaches state-of-the-art performance on ImageNet-1k with fewer parameters and lower FLOPs.

  4. November 2021

    The Florence foundation model

    In November 2021, he is first author of Florence, a computer vision foundation model that extends visual representations from scenes to objects, from images to videos, and from RGB to captions and depth. The model achieves new state-of-the-art results on the majority of 44 representative benchmarks. He is a Principal Research Manager on Microsoft Research's Project Florence-VL team.

  5. 2022

    GLIP at CVPR 2022

    In 2022, he coauthors the grounded language-image pre-training model GLIP, published at CVPR 2022 with Liunian Harold Li, Pengchuan Zhang, Jianfeng Gao and others. GLIP unifies object detection and phrase grounding, and it reaches 49.8 AP on COCO when evaluated directly, without seeing COCO images during pre-training.

  6. April 2024

    The Phi-3 technical report

    In April 2024, he is among the authors of the Phi-3 technical report. It introduces phi-3-mini, a 3.8 billion parameter language model trained on 3.3 trillion tokens that is small enough to deploy on a phone.

  7. June 2024

    Florence-2 at CVPR 2024

    In June 2024, he is the last of nine authors on Florence-2, published at CVPR 2024. The vision foundation model uses a unified, prompt-based representation, and the team co-develops the FLD-5B dataset of 5.4 billion visual annotations on 126 million images.

  8. As of October 2026

    Research manager at Meta GenAI

    As of October 2026, he is a Research Scientist Manager in the GenAI group at Meta. His work centers on multi-modal foundation models, neural image synthesis and editing, and computational photography.

Career

  1. NowMetaResearch Scientist Manager
  2. From 2021Microsoft ResearchPrincipal Research Manager
  3. From 2009Microsoft Research AsiaSenior Research Manager

Education

  1. 2009Hong Kong University of Science and TechnologyPhD
  2. 2005Tsinghua UniversityMS

From the biography.

  1. Google alumni have founded or led 133 other organizations, more than any other AI labIncludes Lu Yuan. We estimate that alumni of 17 major AI labs went on to found or lead 350 other organizations, including 226 they founded, and we trace 472 career moves directly between 23 labs since 2010.
  2. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes Meta. We estimate where the people shaping AI worked when their most-cited papers appeared: the company share was 31% in 2020–2024 and 8% in 1970–1974, and Google led all organizations in 2020–2024.
  3. Companies supply the most prior-work hires at 7 of the 9 frontier AI labsIncludes Meta. Our estimates of the organizations where 550 people worked directly before joining 9 frontier AI labs across 620 career transitions, comparing the main feeder organizations per lab.
  4. Of 943 people shaping AI who worked at Google, 446 work there nowIncludes Meta. Our estimates of where the 943 people shaping AI who worked at Google and Google DeepMind work now: 446 at Google and 452 across 278 other organizations.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Lu Yuan yet. Send us a free-licensed photo.

  1. Lu Yuan | DeepAI deepai.org. Checked 21 October 2026.
  2. [2103.15808] CvT: Introducing Convolutions to Vision Transformers arxiv.org. Checked 21 October 2026.
  3. [2111.11432] Florence: A New Foundation Model for Computer Vision arxiv.org. Checked 21 October 2026.
  4. Project Florence-VL - Microsoft Research: People microsoft.com. Checked 21 October 2026.
  5. [2112.03857] Grounded Language-Image Pre-training arxiv.org. Checked 21 October 2026.
  6. [2404.14219] Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone arxiv.org. Checked 21 October 2026.
  7. Florence-2: Advancing a Unified Representation for a Variety of Vision Tasks - Microsoft Research microsoft.com. Checked 21 October 2026.
  8. [2311.06242] Florence-2: Advancing a Unified Representation for a Variety of Vision Tasks arxiv.org. Checked 21 October 2026.
  9. Lu Yuan - Google Scholar scholar.google.com. Checked 21 October 2026.