Face Recognition Module
Face recognition module
Using people's face recognition to identify their identity is one of the common methods today, which has become one of the attractive methods due to its simplicity and no contact of people with devices, as well as the possibility of recognizing people while moving. This method can be used in different ways, as:
- Using face recognition in movement and time and attendance control devices
- Use in mobile phone applications for different purposes
- Face recognition to enter software systems such as login in Windows
- Use in smart applets for authentication on smart cards
- Use in remote and electronic authentication services (e-KYC)
- And…
Using the knowledge of its artificial intelligence and machine learning specialists and nearly two decades of experience in providing smart remote authentication solutions based on biometrics and access control, Sepid System Company has developed its face recognition algorithm in two forms: services and modules. In this face recognition system, all the necessary parts, such as locating the face in the image (face detection) are included, and face recognition is done accurately and fast. Also, the Sepid face detection algorithm remains efficient in different light conditions, viewing angles and covering conditions.
Sepid face detection service and module
Sepid System has designed and implemented one of the most efficient algorithms for face recognition using modern machine learning methods, esp. deep learning. This algorithm receives a face from an image file (photo) or video stream, and after recognizing the location of the face, normalizing the image, and removing noise, matches the face with the reference photo (such as the person's photo available in the system or the photo on the national card that has been requested from the civil registration organization).
The face detection and tracking capability in online video has made it possible to use this service as an remote authentication solution. Face detection in live videos received from the camera is available both on the client side (web or mobile applications) and on the server side. After detecting the face, it is possible to recognize the face with the face recognition algorithm. The Sepid face recognition algorithm works in two ways: face verification or face identification.
The Sepid face recognition is available as an online service (by calling the relevant APIs from the cloud authentication system of this company) and a module (in the form of an SDK) for local use.
Sepid face detection capabilities
Some of the features of the face recognition system are as follows:
- Presenting as services (with API) and modules (with SDK)
- Using the service and module in different applications and by different developers
- Very low detection error (less than half a percent on LFW reference data)
- Adjusting the sensitivity level of the algorithm and the level of strictness in identification
- High detection speed
Sepid face recognition service and module customers
Any group that needs face-based authentication of people in their products or solutions can use the Sepid face recognition service or module. Some of them are:
- Banks and financial institutions
- Stock exchange organization and related businesses such as brokerages
- Insurance and related companies
- Mobile phone operators
- In-person authentication offices such as government counter and police+10 offices
- Companies and startups providing solutions and products