Exadel, a Silicon Valley-based software consultancy and engineering firm, has updated its CompreFace open source facial recognition service with a slew of new features aimed at improving the customer experience and broadening its integration with other systems.
CompreFace, a face biometrics app that uses deep neural networks, released an initial version in November 2020 for a variety of industries including retail, hospitality, and travel. Exadel released version 1.0 to improve the user experience with a new user interface (UI), no-code options, ease of deployment, and new integrations.
The updated user interface manages subjects and images to begin facial collection without the use of scripts or code. Exadel claims that it also aids in the resolution of facial recognition issues by allowing users to check for correct images saved in each subject.
CompreFace now incorporates facial recognition from closed caption television (CCTV) managed by Home Assistant, an open-source home automation system. With the open source library Double Take, CompreFace developers can now write integrations that allow CCTV facial recognition without writing any custom code. Frigate, a network video recorder with real-time local object detection, is also linked to CompreFace for facial recognition. The file server system unRAID can also be integrated using the CompreFace and Double Take templates.
Serhii Pospielov, head of the AI practice at Exadel, said, “With the support of the developer community, Exadel has continued to improved the CompreFace user experience. As AI practices continue to rise, we are seeing more use cases emerge for facial recognition. This open-source service allows anyone to enter the facial recognition market and access highly reliable services without needing deep coding or machine learning experience.”