Help Center> > Service Overview >What Is Face Recognition

What Is Face Recognition

Face Recognition is an intelligent service that uses computers to process, analyze, and understand facial images based on human facial features. It provides services through open Application Programming Interfaces (APIs). You can obtain the face recognition results by accessing and calling APIs in real time. It recognizes and compares faces automatically and provides you with the similarity degrees, thereby improving service efficiency.

Currently, Face Recognition provides the following sub-services:

  1. Face Detection
  2. Face Verification
  3. Face Retrieval
  4. Face LiveDetect
  5. Face Label

Face Detection

Face Detection accurately locates faces in an image and identifies the size of each face. With this sub-service, you can detect front and side faces with different tilts in the image. It is a basic service in the face recognition field and applies to various scenarios such as security protection, electronic identity, and criminal investigation of public security.
Figure 1 An example of Face Detection

Face Verification

Face Verification returns the similarity degree between two faces in two images by comparing the facial features. If the two images contain multiple faces, the system compares the similarity between the largest faces in the two images.
Figure 2 An example of Face Verification

Face Retrieval

Face Retrieval provides APIs for operating face sets. You can use specific APIs to create a face set, add facial images to the face set, search for N facial images that are most similar to the input one, delete facial images you no longer use from the face set, and delete the face set you created.

Face Retrieval is perfect for enterprise and residential security management, criminal investigation of public security, and other scenarios. However, it cannot be used for preventing photo recapture.
Figure 3 An example of Face Retrieval

Face LiveDetect

Face LiveDetect detects a living figure in a video and the sequence of the figure's actions. If multiple faces appear, the largest face is selected.

Face Label

Labels and clusters faces with the combined functions of face detection, facial gesture recognition, facial and figure feature recognition, and face retrieval algorithms. Users benefit from the ability to retrieve faces by label in real time.

This sub-service is applicable to scenarios with relatively fixed members, such as home monitoring and real-time, automatic recognition of unfamiliar faces in enterprises' monitoring videos. It provides the following capabilities:

  1. Member management: Includes family member input, querying, and deletion functions. The input member images will be provided to user applications as the member labels.
  2. Facial image labeling: Images or videos captured by cameras are dumped to the Object Storage Service (OBS) by user applications and then connected to this service through the terminal-side report interface to implement automatic image labeling. In addition, the label data can be associated with the videos from which the images are generated.
  3. Label-based retrieval: Includes label ID list querying based on face sets and associated data querying based on label IDs.