Function Description
Last updated
Last updated
Excellent aptitude (change of light, change of angle, faces of different age groups) is important for face detection and tracking technology. Our face detection and tracking technology is characterized by its ability to quickly and accurately detect the position of a face in milliseconds for multiple faces in a still or video image on a mobile device or computer. Based on AI deep running technology, it can detect faces in profile, partially hidden faces, blurred face capture photos, and other less-than-perfect facial information, enabling face recognition in a variety of real-world scenarios. The tracking technology is based on the face detection technology and adds face tracking and video analysis technology to increase the speed of face recognition and to quickly detect the position of faces in videos.
Liveness detection uses a series of methods to confirm that the individual currently participating in the test is a real living person instead of a figure from an image or a video clip, or a mask-covered real face.
Static liveness detection technology uses deep learning method to learn the characteristics and differences of living and non-living objects that can be observed by human eyes, and to determine whether the face image is from a living body. It can be effectively used to prevent attacks using images, video remakes, and mask models.
Face verification can calculate the similarity of two faces to determine whether they belong to the same person.
Face verification technology uses deep learning method to extract features from a large amount of data, and uses the degree of feature matching to distinguish different faces, which is suitable for application scenarios including identity recognition and similar face query.
Recommended image size：
Distance between eyes should be larger than 30 pixels.
Face database query can be used to quickly search similar face data in a large-scale face database to identify identity, which can be used in different fields such as celebrity search, monitoring and tracking criminals.
The size of the face library supported by different sub-versions is different, please contact us for details.
Recommended image size：
The minimum number of pixels of the face length should not be less than 1/8 of the length of the image short edge.
Distance between eyes should be larger than 30 pixels.
