Applied Recognition receives two face recognition patents
Applied Recognition announced it has received two patents related to its face recognition technology.
The first is from the Chinese State Intellectual Property Office (No. 200880126543.0) entitled “Method, System and Computed Program for Identification and Sharing of Digital Images with Face Signatures”, while the second patent is from the Canadian Intellectual Property Office under the same title (No. CA 2711143).
These patents are similar to patent No. 55755185-1US, which was issued to Applied Recognition by the US Patent Office last year. All three patents share the priority date of December 31, 2007.
The patents cover Applied Recognition’s facial technology which automatically recognizes multiple known faces in photos or videos on a desktop computer or mobile device.
The technology ensures the advanced organization and presentation of photos or videos based on the graphical selection of known faces by selecting thumbnail images of people.
The patents also cover the method of sharing and distributing photos or videos in an automated process between “friends” who are using the same software that enables the invention.
Finally, these patents grant cover the ability to review the results of automatic face detection, eye detection and face recognition methods, as well as to correct any errors resulting from the automated process.
Applied Recognition said it began work on the process of organizing consumer photos using face recognition well before the December 2007 priority date, and claims it is the first company to conceive and develop such an application.
The company’s patents perform the search for faces in unconstrained environments and render the found faces into posed positions, similar to a mug shot database.
The digital face signatures, which are generated from the rendered face, are clustered to group like faces together for the purpose of easy tagging.
By comparing digital signatures of unknown faces with known faces, the system suggests matches to the consumer who can confirm or reject the suggestions, which enables the recognition software to improve over time.
Applied Recognition’s first product, Fotobounce, was independently tested against Google Picasa, Apple iPhoto and Microsoft Photo Gallery, in which it outperformed all three products in the categories of speed, accuracy and efficiency.