Face.com

Face.com was an Israeli technology company established in Tel Aviv in 2009 to develop facial-recognition software for photographs uploaded through web and mobile applications. Its platform was designed to detect and identify faces efficiently at scale, then associate recognized people with tagging and social-network data. The company operated with a small team, reported in reference material as approximately ten employees, but its software was used through consumer applications and an application programming interface rather than through a large physical retail or hardware footprint. The company’s most visible early products were Facebook applications. Photo Finder helped users locate photographs on Facebook in which they or their friends had not yet been tagged, while Photo Tagger supported the bulk tagging of faces appearing across multiple photographs. These applications positioned Face.com as an infrastructure and utility provider for social-photo organization rather than as a conventional consumer-facing social network. The underlying technology was intended to reduce the effort required to identify people in large image collections. Face.com also pursued a developer-platform strategy. In March 2010 it opened an alpha version of a free REST API that enabled third-party developers to incorporate its facial-recognition algorithms and related tagging data into their own applications and services. The API supported recognition and tagging workflows involving Facebook and Twitter, and could also be used with image collections associated with services such as Flickr and Picasa. Face.com offered rate-limited access, developer whitelisting, and premium licensing options. In February 2011 the API moved from alpha to beta, with the stated hourly scanning limit increasing from 200 to 5,000 photographs. Reference material reported that more than 10,000 developers were using the API at that point. Face.com claimed substantial processing scale. By February 2011, its API and Facebook applications were reported to have identified 18 billion faces. A 2011 University of Massachusetts Amherst study comparing part of the company’s algorithm reported a recognition rate of 0.91 for the evaluated portion of the system. Such results and usage claims helped establish Face.com as an early specialist in cloud-based facial recognition for social photography, at a time when image tagging and mobile photo management were rapidly expanding. The company raised a reported total of $5.3 million in venture funding. A $1 million Series A round was announced in February 2009, followed by a $4.3 million Series B round announced on September 27, 2010. The Series B was led by existing investor Rhodium and Russia-based technology company Yandex. Face.com’s independent operation ended in 2012, when Facebook announced on June 18 that it had acquired the company for an undisclosed amount. Contemporary estimates placed the transaction at approximately $55 million to $60 million, although the consideration was not publicly disclosed. Face.com said it would wind down its public APIs and focus on products for Facebook. Its facial-recognition technology continued to influence Facebook’s research and product development, including the DeepFace project. The Face.com domain was later sold to an e-commerce company in December 2015, reflecting the end of the original brand’s standalone web presence.

Face.com emerged in Tel Aviv in 2009 during the rapid growth of online photo sharing and social-network platforms. Its central proposition was that facial recognition could automate the otherwise manual work of finding, identifying, and tagging people in digital photographs. Rather than building a general-purpose social network, Face.com supplied recognition technology through its own Facebook applications and through tools intended for outside developers. The company released Photo Finder and Photo Tagger on Facebook. Photo Finder helped users identify untagged photographs of themselves and friends, while Photo Tagger enabled bulk tagging of faces across multiple images. These applications provided a consumer demonstration of Face.com’s technology and gave the company access to large-scale social-photo use cases. In March 2010 Face.com expanded beyond its own applications by opening an alpha REST API. The service allowed third-party developers to use facial-recognition algorithms and related tagging capabilities in their own products. It supported workflows involving Facebook and Twitter and could be applied to image collections from multiple sources, including Flickr and Picasa. The company offered free access subject to rate limits, as well as whitelisting and premium licensing arrangements. This API strategy made Face.com an early example of a specialized machine-learning company commercializing recognition infrastructure as a service. Face.com announced a $1 million Series A financing in February 2009 and a $4.3 million Series B financing on September 27, 2010. The latter round was led by Rhodium, an earlier investor, together with Yandex. Reference material states that total funding reached $5.3 million. By February 2011, the company reported that its API and Facebook applications had discovered 18 billion faces. Its API moved from alpha to beta that month, increasing the stated hourly limit from 200 photographs to 5,000, and more than 10,000 developers were reported to be using it. The technology also attracted academic evaluation. A 2011 University of Massachusetts Amherst study compared a portion of the Face.com algorithm and reported a recognition rate of 0.91 for the portion examined. The result placed the company among the early commercial facial-recognition specialists whose products combined image analysis with social identity data. On June 18, 2012, Facebook announced the acquisition of Face.com. The purchase price was not disclosed publicly; contemporary estimates cited in reference material placed it between $55 million and $60 million. After the transaction, Face.com announced that its public APIs would be wound down and that the team would focus on new Facebook products. The company’s technology continued within Facebook-related research and development, later associated with the DeepFace project. Face.com therefore ceased to operate as an independent technology brand, although its core engineering work became part of Facebook’s broader facial-recognition efforts. In December 2015, the Face.com domain was reported as having been sold to an e-commerce company.

• 2015

  Face.com domain changes ownership

  [1]Face.comen.wikipedia.org

  The Face.com domain is reported to have been sold to an e-commerce company.

• 2012

  Facebook acquisition announced

  [1]Face.comen.wikipedia.org

  Facebook announces its acquisition of Face.com and the planned closure of Face.com’s public APIs.

• 2011

  API enters beta

  [1]Face.comen.wikipedia.org

  The API moves from alpha to beta, with the stated scanning limit rising to 5,000 photographs per hour and reported adoption exceeding 10,000 developers.

• 2010

  REST API alpha launches

  [1]Face.comen.wikipedia.org

  Face.com opens a free, rate-limited REST API for developers building applications around facial recognition and image tagging.

• 2010

  Series B financing

  [1]Face.comen.wikipedia.org

  The company announces a reported $4.3 million Series B led by Rhodium and Yandex.

• 2009

  Face.com is established in Tel Aviv

  [1]Face.comen.wikipedia.org

  The Israeli facial-recognition technology company is founded and begins developing software for identifying and tagging faces in online photographs.

• 2009

  Series A financing

  [1]Face.comen.wikipedia.org

  Face.com announces a reported $1 million Series A financing round.

An early cloud-based facial-recognition and photo-tagging platform for social networks, developers, and large online image collections.

• Facebook application

  Photo Finder

  Photo Finder was a Facebook application designed to locate photographs in which a user or the user’s friends appeared without being tagged. It demonstrated Face.com’s ability to scan image collections, detect faces, and connect them with social-network identities. The application addressed the practical problem of finding overlooked photographs in large personal archives and helped expose the company’s recognition technology to ordinary Facebook users.

• Facebook application

  Photo Tagger

  Photo Tagger enabled bulk tagging of faces appearing across multiple photographs uploaded to Facebook. Instead of requiring users to label each image individually, the application applied automated face detection and recognition to accelerate the tagging workflow. It was one of Face.com’s principal consumer-facing demonstrations of how machine learning could organize and annotate social photographs.

• Developer platform · 2010

  Face.com REST API

  The Face.com REST API allowed external developers to incorporate the company’s facial-recognition algorithms and tagging functionality into their own applications and services. It supported recognition workflows involving Facebook and Twitter and could be combined with image sources such as Flickr and Picasa. The program began as a free alpha service with rate limits, developer whitelisting, and premium licensing options before moving to beta in 2011.

• 2012

  Join Facebook and wind down the public APIM&A

  Background

  Facebook acquired Face.com to obtain facial-recognition expertise and incorporate the technology into Facebook products.

  Action taken

  Face.com announced that its public APIs would be wound down and that the team would concentrate on new Facebook products.

  Aftermath

  Face.com ceased to operate as an independent company, while its technology continued within Facebook-related research and product development, including the DeepFace project.

  Estimated acquisition value. $55 million-$60 million estimated transaction value (2012 acquisition; estimate reported contemporaneously, while the purchase price was undisclosed)

  Source

• 2011

  Move the API from alpha to betaGeneration change

  Background

  After operating the developer service in alpha, Face.com sought to support broader use and higher-volume image processing.

  Action taken

  The company moved the API to beta and increased the stated hourly scanning limit from 200 to 5,000 photographs.

  Aftermath

  Reference material reported that more than 10,000 developers were using the API during this period.

  Source

• 2010

  Open the REST API to third-party developersProduct launch

  Background

  Face.com sought to extend its facial-recognition technology beyond its own Facebook applications and reach developers building services around online photographs.

  Action taken

  The company opened an alpha version of a free REST API with rate limits, whitelisting options, and premium licensing possibilities.

  Aftermath

  The API later moved to beta, increased its processing limits, and was reported to have attracted more than 10,000 developers.

  Source

1. 1Face.comen.wikipedia.org