Clarifai Inc. is an artificial intelligence (AI) company that specializes in computer vision and uses machine learning and deep neural networks to identify and analyze images and videos. The company offers its solution via API, mobile SDK, and on-premise solutions.

Clarifai
Private
IndustryComputer Vision
Founded2013 (New York City)
Headquarters,
U.S.
Area served
Worldwide
Key people
Matthew Zeiler (Founder & CEO)
Number of employees
90+
Websiteclarifai.com

Clarifai is headquartered in New York City with two satellite offices in San Francisco and Washington D.C.

HistoryEdit

Clarifai was founded in 2013 by Matthew Zeiler, a Ph.D. student who recently placed in the top 5 spots of the 2013 ImageNet Challenge.[1][2][3] Initially, the company offered free and paid versions of image and video recognition via their API and a consumer-facing iPhone app called Forevery.[4][5] In 2016, Clarifai released version 2 of their API, adding custom training and visual search to its platform.[6][7]

In 2017 the company moved all research work to a San Francisco office and all government-related endeavors to an office in Washington D.C.[8] Later that year, the company announced a mobile SDK, which allowed users to run their platform without an internet connection.[9][10][11] In 2018 Clarifai released an on-premise solution.[12]

FundingEdit

In 2015, Clarifai raised $10 million in its Series A funding round, led by Union Square Ventures (USV).[13] After the 2016 launch of their v2 API, Menlo Ventures led their $30 million Series B round, with participation from USV, Lux Capital, and Osage University Partners.[14]

TechnologyEdit

The core of Clarifai's technology is based on convolutional neural networks, which Zeiler focused on for his PhD work.[15] It is a process which enables a computer to learn from data examples and draw its own conclusions, giving applications the ability to predict correct tags for images or videos.

The platform includes the ability to moderate content, perform visual search, visual similarity, and organize media collections. It has pre-built recognition models that can identify a specific set of concepts like food or travel, NSFW, and its general model which can identify a range of concepts including objects, ideas, and emotion.[16] It also has the ability to create custom models which can identify other arbitrary objects such as cars or breeds of dogs.[17] The 2018 Model 1.5 with machine-labeled datasets claims to recognize up to 11,000 concepts from object detection, as well as things like mood or theme.[18]

Military workEdit

In 2018 Zeiler disclosed that the company was a participant in Project Maven, a US Department of Defense AI program.[19] The disclosure, in a blog post on the company website, came after a Wired story reported that a former employee had filed a wrongful termination suit. The suit alleged that she was dismissed for requesting that Clarifai disclose a 2017 server compromise to the Pentagon and other customers.[20] Zeiler asserted that the breach involved only "an isolated research server", and that customers were notified following an external security audit.[19]

According to employees Clarifai intends to 'contribute to autonomous weapons'.[21]

ReferencesEdit

  1. ^ Tilley, Aaron. "Every AI Powerhouse Wanted This Whiz Kid. He's Taking Them on Instead". Forbes. Retrieved January 31, 2018.
  2. ^ "How Matthew Zeiler Created 'Clarifai' – The Image And Video Recognition API For Building Smart Apps". CrazyEngineers. Retrieved January 31, 2018.
  3. ^ Gershgorn, Dave. "The data that transformed AI research—and possibly the world". Quartz. Retrieved May 22, 2018.
  4. ^ "The AI Startup Google Should Probably Snatch Up Fast". WIRED. Retrieved May 22, 2018.
  5. ^ Hernandez, Daniela. "This app will actually make you look at your millions of forgotten photos". Splinter. Retrieved June 7, 2018.
  6. ^ "Clarifai Wants You to Correct AI's Biggest Gaffes". WIRED. Retrieved May 22, 2018.
  7. ^ "Deep learning startup Clarifai raises $30 million". VentureBeat. October 25, 2016. Retrieved May 22, 2018.
  8. ^ "Visual search startup Clarifai beefs up its AI team with a new San Francisco office – TechCrunch". techcrunch.com. Retrieved May 22, 2018.
  9. ^ "AI company Clarifai says its new face recognition technology won't be racist". Fast Company. May 1, 2017. Retrieved May 22, 2018.
  10. ^ "How Machine-Learning AI Is Going To Make Your Phone Even Smarter". Fast Company. July 12, 2017. Retrieved May 22, 2018.
  11. ^ "Clarifai launches SDK for training AI on your iPhone". VentureBeat. July 12, 2017. Retrieved May 22, 2018.
  12. ^ "Clarifai rolls out an on-premise visual search tool for businesses – TechCrunch". techcrunch.com. Retrieved May 22, 2018.
  13. ^ "Deep-learning startup Clarifai gets $10M to make business hires, add features". VentureBeat. April 28, 2015. Retrieved May 22, 2018.
  14. ^ "Clarifai raises $30M to give developers visual search capabilities – TechCrunch". techcrunch.com. Retrieved May 22, 2018.
  15. ^ Zeiler, Matthew D.; Fergus, Rob (November 12, 2013). "Visualizing and Understanding Convolutional Networks". arXiv:1311.2901 [cs.CV].
  16. ^ "Models". Clarifai. Retrieved May 22, 2018.
  17. ^ "Clarifai Raises $30 Million to Innovate Artificial Intelligence for Everyone". Techvibes. Retrieved May 22, 2018.
  18. ^ "Clarifai improves General Model 1.5 with machine-labeled datasets". VentureBeat. October 4, 2018. Retrieved November 19, 2018.
  19. ^ a b Zeiler, Matthew (June 14, 2018). "Why We're a Part of Project Maven". Clarifai. Retrieved July 8, 2018.
  20. ^ Simonite, Tom (June 12, 2018). "Startup Working on Contentious Pentagon AI Project was Hacked". Wired. Retrieved July 8, 2018.
  21. ^ Is Ethical A.I. Even Possible? The New York Times, 2019