Matroid (company)

Matroid
Industry Computer software
Founded 2016 (2016)
Founder Reza Zadeh
Headquarters Palo Alto, California
Website matroid.com

Matroid is a Computer Vision company that gives typical computer users the ability to scan video. Matroid was founded in 2016 in Palo Alto, California.

Product

The Matroid Product is a studio for creating and using detectors that scan video for faces, actions, places, and other objects. Example applications include determining when former President Barack Obama appears on TV or how often a man holding a gun is recorded in security footage. Users can create custom detectors to find specific people or objects, or they can pick from a library of pre-programmed detectors designed by other users of the platform.[1][2][3][4]

Matroid is initially focused on analyzing television appearances and scanning surveillance video. In the first case, a firm may want to track which political candidates got more TV time or which brand of car appears more often on a particular show or television network.

Product features include a video player with computer at its core, a tight feedback loop for detector iteration, an extensive API, stream monitoring, and work-sharing amongst users in a library of public detectors.

Microsoft and Google presented at the 2017 Scaled Machine Learning Conference held at Stanford University.

Scaled Machine Learning Conference

Matroid holds the Scaled Machine Learning Conference (ScaledML) every year. The previous two conferences were held on the Stanford University campus.[5]

At ScaledML 2016, Microsoft announced "Open Mind", Microsoft's Visual Studio-like suite for machine learning.[6][7][8] At the 2017 conference, Google announced the Tensor Processor Unit to speed up machine learning operations in Google Cloud Platform.[9]

Funding

Matroid raised $3.5 million dollars in funding from New Enterprise Associates in 2016, and a further $10 million from Intel in 2017.[10][11][12]

Open Source

Matroid uses TensorFlow and Kubernetes. As an active member of the TensorFlow open-source community, Matroid has held TensorFlow tutorials[13] and is currently writing a book on TensorFlow for Deep Learning with O'Reilly Media.[14] For its contributions to open-source machine learning, Matroid was awarded a best paper award at KDD 2016.[15]

References

  1. ↑ "This AI Company Can Tell You What and Who Appears in Your Videos". Bloomberg.com. 2017-03-25. Retrieved 2017-04-13.
  2. ↑ Martin, Scott (2017-03-27). "A Life's Ambition, Matroid Launches". Wall Street Journal. ISSN 0099-9660. Retrieved 2017-04-13.
  3. ↑ Mannes, John. "Matroid can watch videos and detect anything within them". TechCrunch. Retrieved 2017-04-13.
  4. ↑ "The Vast, Secretive Face Database That Could Instantly ID You In A Crowd". Fast Company. 2017-03-30. Retrieved 2017-04-23.
  5. ↑ "Scaled Machine Learning". scaledml.org. Retrieved 2017-04-13.
  6. ↑ Foley, Mary Jo. "Open Mind: Microsoft's Visual Studio-like suite for machine learning | ZDNet". ZDNet. Retrieved 2017-04-13.
  7. ↑ "Microsoft's Open Mind Studio is the "Visual Studio" for Machine Learning - Artificial Intelligence Online". Artificial Intelligence Online. 2016-08-17. Retrieved 2017-04-13.
  8. ↑ "Microsoft's Qi Lu reveals Open Mind, the "Visual Studio for Machine Learning" | On MSFT". On MSFT. 2016-08-17. Retrieved 2017-04-13.
  9. ↑ Dean, Jeff. "Scaling Machine Learning with TensorFlow" (PDF).
  10. ↑ Mannes, John. "Matroid picks up $10M Series A to automate video stream monitoring". TechCrunch. Retrieved 2017-09-22.
  11. ↑ Martin, Scott (2017-03-27). "A Life's Ambition, Matroid Launches". Wall Street Journal. ISSN 0099-9660. Retrieved 2017-04-13.
  12. ↑ "If you can't impose self-discipline, you can't be better off as a private company".
  13. ↑ "Scaled Machine Learning". scaledml.org. Retrieved 2017-04-13.
  14. ↑ Ramsundar, Bharath. "TensorFlow for Deep Learning". O'Reilly Media.
  15. ↑ News, SIGKDD. "SIGKDD Awards : 2016 SIGKDD Best Paper Award Winners". www.kdd.org. Retrieved 2017-04-13.
This article is issued from Wikipedia. The text is licensed under Creative Commons - Attribution - Sharealike. Additional terms may apply for the media files.