TellWorldAI World Archive

Home›Companies

OrganizationResearch lab

Microsoft Research Asia

Microsoft Research Asia is Microsoft's research laboratory in the Asia-Pacific region, founded in Beijing in November 1998. It works on artificial intelligence, human-computer interaction and related fields, and its researchers have included the first author of the ResNet paper.

Founded 1998Website microsoft.com/en-us/research/lab/microsoft-research-asia Era Statistical learning

What it is

Microsoft Research Asia describes itself as Microsoft's research arm in the Asia-Pacific region and says it was established in 1998.[7] Its main base is in Beijing, where it has more than 300 researchers and developers, together with roughly 300 visiting scientists and students; it also runs a satellite office in Shanghai.[1] The laboratory's own site lists research labs and personnel in Beijing, Shanghai, Vancouver, Tokyo, Singapore and Hong Kong.[7]

Where it came from

The lab was founded in Beijing in November 1998.[1] Its first director was Kai-Fu Lee, who led it from 1998 to 2000, when it was called Microsoft Research China; it was later renamed Microsoft Research Asia.[2] That sequence — a China-based lab that grew into a region-wide one — is reflected in the name it carries today.

What it works on

A listing of the lab's research areas includes natural user interfaces, multimedia, data-intensive computing, search and online advertising, natural language processing and computer science fundamentals.[1] The areas highlighted on its own site are grouped differently: AI Foundation & Systems, Societal AI & HCI, Spatial Intelligence & Immersive AI, and Wellbeing & Neuroscience.[6] The two lists taken together show a lab that began with interfaces and media and has moved a large part of its attention toward artificial intelligence and the study of how people interact with computing systems.

People and results

Kaiming He worked as a researcher at Microsoft Research Asia from 2011 to 2016.[3] In December 2015 he was first author, with Xiangyu Zhang, Shaoqing Ren and Jian Sun, of the paper introducing residual networks, or ResNet.[4] The idea behind a residual network is that blocks of layers learn a correction to their input rather than the full mapping from input to output, which makes very deep networks easier to train. The networks described in the paper were up to 152 layers deep, and they won the ILSVRC 2015 image-recognition competition.[4] Depth of this kind had been difficult to train before, so the result became a widely used reference point for later work on deep networks.

Place in the organisation

The Beijing lab sits inside Microsoft's Asia-Pacific Research and Development Group, which also has campuses in Suzhou, Shenzhen, Tokyo and Taipei.[1][5] Separate from that structure, a Microsoft Research Asia–Tokyo laboratory opened in November 2024.[5] Read together, these facts describe a network rather than a single building: a founding site in Beijing, a satellite in Shanghai, branches elsewhere in China and Japan, and staff and labs listed in Vancouver, Singapore and Hong Kong as well.[1][5][7]

CapabilitiesComputer vision

TechniquesDeep learning

Ecosystem & regionChina

HistoryResearch history

20181

  1. MilestoneMachine reading comprehension surpasses human baseline on SQuAD

    In January 2018, Microsoft Research Asia's R-NET and then an Alibaba model scored above the human baseline on SQuAD, a machine reading comprehension benchmark — the first time a system had exceeded human exact-match performance there.

20161

  1. MilestoneResNet wins the CVPR 2016 Best Paper Award

    At CVPR 2016 in Las Vegas, the Best Paper Award went to "Deep Residual Learning for Image Recognition" by Kaiming He, Xiangyu Zhang, Shaoqing Ren and Jian Sun. Google Scholar later reported it as the most cited research paper over five years.

Notes and sources

  1. Microsoft Research Asiahttps://en.wikipedia.org/wiki/Microsoft_Research_Asia
  2. Kai-Fu Leehttps://en.wikipedia.org/wiki/Kai-Fu_Lee
  3. Kaiming Hehttps://en.wikipedia.org/wiki/Kaiming_He
  4. Deep Residual Learning for Image Recognitionhttps://arxiv.org/abs/1512.03385
  5. Microsoft Researchhttps://en.wikipedia.org/wiki/Microsoft_Research
  6. Microsoft Research Asiahttps://www.microsoft.com/en-us/research/lab/microsoft-research-asia/
  7. Microsoft Research Asia - About Ushttps://www.microsoft.com/en-us/research/lab/microsoft-research-asia/about-us/

Cite this page: Microsoft Research Asia. TellWorld, https://www.tellworld.net/companies/microsoft-research-asia. Updated 2026-11-08