PeopleofSuperintelligence

Researcher

Lin Ma

Lin Ma is a computer-vision researcher whose deep learning work connects images and video with language. A graduate of Harbin Institute of Technology, he completed a Ph.D. in electronic engineering at the Chinese University of Hong Kong in 2013, then worked at Huawei Noah's Ark Lab in Hong Kong, where he coauthored multimodal convolutional neural networks for matching image and sentence. As a Principal Researcher at Tencent AI Lab from 2016 to 2020, he coauthored Reconstruction Network for Video Captioning in 2018, and his 2019 paper Image Deformation Meta-Networks for One-Shot Learning was a CVPR best paper finalist. He is a researcher at Meituan in Beijing, where he coauthored the Flash-VL 2B vision-language model in 2025.

Now at
Meituan
Badges

Timeline

  1. 2006–2008

    Degrees at Harbin Institute of Technology

    He earns his B.E. in 2006 and his M.E. in 2008, both in computer science, at Harbin Institute of Technology. He is a research intern at Microsoft Research Asia from October 2007 to March 2008.

  2. 2008–2013

    A Ph.D. at CUHK

    He is a research assistant in the Department of Electronic Engineering at the Chinese University of Hong Kong (CUHK) from November 2008 to July 2009. He completes his Ph.D. in that department in 2013.

  3. 2008–2012

    Microsoft Research Asia fellowship

    He wins the best paper award at the Pacific-Rim Conference on Multimedia (PCM) in 2008. He receives the Microsoft Research Asia fellowship in 2011 and is a finalist for the HKIS young scientist award in engineering science in 2012.

  4. 2013–2016

    Researcher at Huawei Noah's Ark Lab

    He is a researcher at Huawei Noah's Ark Lab in Hong Kong from August 2013 to September 2016. In 2015, he publishes multimodal convolutional neural networks (m-CNNs) for matching image and sentence at ICCV, and the networks reach state-of-the-art results on Flickr30K and Microsoft COCO retrieval.

  5. 2016–2020

    Principal Researcher at Tencent AI Lab

    He is a Principal Researcher at Tencent AI Lab in Shenzhen from September 2016 to June 2020. In 2018, he coauthors Reconstruction Network for Video Captioning, a CVPR paper whose encoder-decoder-reconstructor model uses both the video-to-sentence and sentence-to-video flows.

  6. 2019

    A CVPR best paper finalist

    His paper Image Deformation Meta-Networks for One-Shot Learning is one of 50 best paper finalists among the 1294 papers accepted to CVPR in 2019. The method pairs a meta-learner with an image deformation sub-network that produces extra training examples.

  7. 2022–2023

    Winning computer-vision challenges

    In 2022, his team takes first place in both the Scan2Cap and ScanRefer challenges at the 2nd Workshop on Language for 3D Scenes. In 2023, it wins first place in the Naturalistic Driving Action Recognition track of the 7th AI City Challenge Workshop during CVPR, and second place in the SoccerNet Action Spotting and Multiple Object Tracking challenges.

  8. As of October 2026

    Researcher at Meituan

    He is a researcher at Meituan in Beijing. In May 2025, he coauthors Flash-VL 2B, a vision-language model built for ultra-low latency and high throughput and tested on 11 standard benchmarks, and in August 2025 he is invited as an Area Chair for ICLR 2026.

Career

  1. NowMeituanResearcher
  2. 2016–2020TencentPrincipal Researcher
  3. 2013–2016HuaweiResearcher
  4. 2008–2009Chinese University of Hong KongResearch Assistant
  5. 2007–2008Microsoft Research AsiaResearch Intern

Education

  1. 2013Chinese University of Hong KongPh.D.
  2. 2008Harbin Institute of TechnologyM.E.
  3. 2006Harbin Institute of TechnologyB.E.

From the biography.

  1. Google alumni have founded or led 133 other organizations, more than any other AI labIncludes Lin Ma. 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. 23% of people shaping AI who left a university for a company later returned to a universityIncludes Lin Ma. We estimate that 468 of the 2,011 people shaping AI who left a university for a company (23%) later returned to a university after a median of 5 years, and 163 of 829 (20%) made the reverse round trip.
  3. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other universityIncludes Lin Ma. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other university. We compare doctorate and 5+ year employment links, career timing and current workplaces for 187 major-prize and 1,111 best-paper winners.
  4. Among employers with 30+ full-time cases, OpenAI leads primary technical role H-1B pay at $310,000, and Amazon, Google and Microsoft file 57% of all casesIncludes Microsoft. Across 440,669 certified US H-1B filings from 44 employers (FY2020–FY2026), OpenAI leads primary technical role base pay at $310,000 and xAI leads overall base pay at $290,000.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

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

  1. Lin Ma forestlinma.com. Checked 21 October 2026.
  2. 20180606-16 马林：Image/video Captioning - 好文作者面授招 - VALSE valser.org. Checked 21 October 2026.
  3. Multimodal Convolutional Neural Networks for Matching Image and Sentence arxiv.org. Checked 21 October 2026.
  4. Reconstruction Network for Video Captioning (CVPR 2018 Open Access Repository) openaccess.thecvf.com. Checked 21 October 2026.
  5. Lin Ma | Awards forestlinma.com. Checked 21 October 2026.
  6. Image Deformation Meta-Networks for One-Shot Learning (CVPR 2019 Open Access Repository) openaccess.thecvf.com. Checked 21 October 2026.
  7. Lin Ma | Publications forestlinma.com. Checked 21 October 2026.
  8. Flash-VL 2B: Optimizing Vision-Language Model Performance for Ultra-Low Latency and High Throughput arxiv.org. Checked 21 October 2026.