DeepGlint Face Recognition System

DeepGlint Face Recognition System

DeepGlint face recognition system enables human detection and face recognition in vehicles at security gate entrances, based on on-premise security big data set. In the field of security monitoring, the identification of dynamic faces and static faces is the main demand of the current market. By solving the face recognition and comparison problems in non-contact and non-cooperating monitoring scenarios, face recognition has become one of the most important bio-metric identification methods, which has great application prospects and commercial value. With a large amount of data in deep learning technology and security monitoring, Deepglint has developed a leading international algorithm for security monitoring. Its face recognition system, including two subsystems: face recognition and static large library face comparison. It can be widely used in public security, airports, customs, key places access control and other fields.

*Please note that member solutions are often customizable to meet the needs of individual enterprise end users.



  • High face recognition contrast accuracy
    Face (1 vs1) mismatch rate is less than one billionth. Dynamic face recognition comparison accuracy is above 95%.
  • Strong face detection and capture ability
    Support simultaneous detection of hundreds of faces in the picture. Support for detecting low-profile faces. Support for detecting high occlusion faces.
  • Support multiple face attributes
    Support gender, age, glasses, hat and other face attributes.



China (PRC) Intel® Xeon Scalable On-premise (Private Cloud, Other) Caffe Cross-Industry Finance and Insurance Government Retail Models cant be re-trained - Inference only Linux Intel® Math Kernel Library (Intel® MKL) Other ResNet50 Facial Detection/Recognition/Classification Image/Object Detection/Recognition/Classification Smart City Deep Learning