Introduction to privatized face recognition deployment

Face recognition privatisation deployment (also known as facial recognition localization deployment) provide access to face detection, face alignment, face, live detection, detecting head and shoulders, human attributes such as face recognition foundation ability, allows users to quickly integrate face recognition ability, expand its business system, save a lot of research and development and the time cost, quick hug AI era.

Face recognition privatized deployment applies

The privatized face recognition can be deployed locally and run on a LAN without network restrictions. Compared with public cloud API calls, the privatized face recognition provides higher security and flexibility, and is more suitable for demanding scenarios.

Privatized face recognition deployment system architecture

Privatized face recognition deployment high expansion, high performance architecture diagram,

Provides REST interfaces for seamless integration.

Support millions and millions of bottom library,

It can run on a CPU or GPU.

Terminology Tips

  • What is face detection?

Given a photo, determine whether it contains a face, if it does, return the position, size, and attribute analysis results of the face.

  • What is face comparison?

Two photos of faces are calculated for similarity to determine if they are the same person and given a similarity score. It can be used in scenarios such as real identity authentication and human-witness authentication.

  • What is face retrieval?

Given a photo, compare it to multiple faces in the face library to find the most similar one. According to the given face and face library matching degree, return matching degree, that is, face retrieval, also known as 1: N face retrieval.

  • What is a silent living body?

Based on the portrait flaws (mole lines, imaging malformations, etc.) in the picture, it can determine whether the target is alive or not without the user’s cooperation to complete the specified action, which can effectively prevent cheating behaviors such as second retakes. It can be used for witness verification, self-service machine services, etc.

  • What are face attributes?

Face attributes for a series of face related attributes analysis, including age, gender, face Angle and other attributes.

  • What are anthropogenic attributes?

Detect all the human bodies in the image and identify more than 20 categories of attribute information of human bodies, including gender and age, headwear, hair color and length, clothing category, clothing color and wearables, etc.

  • What is face database management?

In face retrieval, the need for face classification and input face base, face retrieval can be in the specified one or more face database search.

Read more: ai.minivision.cn/#/solution/…