Face Recognition
Import or identify visitors with frame-by-frame detection in a dynamic retail setup
Our face recognition is well suited for retail businesses, infrastructure sites with critical access control needs, HoReCa, and gambling establishments.
NSoft Vision face recognition is capable of identifying and verifying a person from a visual frame of a video source. Our in-house developed neural network algorithms are based on machine learning techniques, which emulate how the human brain operates. Face recognition was trained on a huge dataset with labeled faces to map a face to a numerical vector representation.
We offer high-quality predictions supporting recognition with the camera angle, different lighting, hairstyle, and facial expression.
We offer high-quality predictions supporting recognition with the camera angle, different lighting, hairstyle, and facial expression.
Visitors are registered by the system automatically with anonymized identification and can be imported by uploading a picture, snapshot, or photo-fit. Each identity can be further managed or compared to a list of similar identities and merged in case there was a low-quality face appearance.