The modern application of face detection technology with high quality cameras is viable option for authentication as well as identification.
We developed the portfolio that helps admins of a store to analyze his store activities.Analytics uses images captured by IP Cameras situated at various places of a store. Face Detection is done for each of the captured images and Face Comparison is done for detected faces for assigning an identity to the face. Main features of the application are Store View, total pedestrians count values, Customer and Employee Heat Map showing where customer/employee spent more time,Average Visitors of the day, week and month,Gender and Age wise counts, Holiday/Busiest time customer count and trend comparison.
In Face Recognition application, we used Openface algorithm with siamese network ,one module for face detection and another for Face Search.
From face detection API our application took the co-ordinate of the face bounding box, gender, age, face quality, blur and head pose of the face for each face present in the input image.The Face token/Id of the detected faces are stored in a collection at Face++ server (Using another set of APIs for creation of this collection and adding faces to it).Finally,the newly detected faces are compared with this collection using face token, the search API is expected to return most matching faces with their token and match confidence value.