Rapid and correct multi-pose real-time detection
Light normalization pretreatment and shadow elimination
Exact matching under remote distance and low resolution
Accurate analysis of face after long time span
Evaluation and analysis of face age and gender
Facial Feature Detection based Algorithm
Sensingtech face recognition algorithm adopts multi-angle face detection and positioning. The system has the face recognition and authentication capacity at a level of over 99%, with the recognition speed of 1s to 2s. On occasions of flow density and appearance of multiple human faces in an image, the system will make rapid dynamic detection and positioning of human faces of objects that are moving and not controlled.
Deep Learning Model
Deep learning model is recognized as the most advanced face recognition algorithm worldwide. Sensingtech dynamic face recognition technology works on the basis of convolutional neural network’s deep monitoring and learning model and million-level face data set. With high system recognition rate and response rate, Sensingtech keeps the leading position worldwide in face recognition technology. It has been rated as no.1 for two consecutive years in TRECVID (TREC Video Retrieval Evaluation) evaluation.
Sensingtech has developed a variety of products, and praise by the industry.
Control server of static face and historical monitoring video
Sensingtech application platform of static face and historical video