Skip to main content
Version: 1.1.0

Architecture

img.png

Image API consists of the following services:

  • gateway: the service is responsible for getting access to Image API and interacting with the processing services listed below;
  • body-detector: the service is designed to detect bodies in an image. The detection result is the coordinates of the bounding box (bbox) around the detected body;
  • face-detector-face-fitter: the service is used to determine the anthropometric points and the head rotation angles. The returned result is a set of 21 anthropometric points and the values of head rotation angles (yaw, pitch, roll);
  • emotion-estimator: the service estimates a person's emotions from a face image. The result is the degree of manifestation of each estimated emotion;
  • age-estimator: the service estimates a person’s age from a face image. The estimation result is the numerical value of the person's age;
  • gender-estimator: the service is used to determine a person’s gender from a face image;
  • face-detector-liveness-estimator: the service is used to detect if a person in the image is real or fake. The result is a Real/Fake verdict with a numerical value of confidence. This service has three modules: liveness-anti-spoofing, quality-liveness-anti-spoofing, face-detector-liveness-estimator. Modules liveness-anti-spoofing and face-detector-liveness-estimator use different liveness algorithms. In quality-liveness-anti-spoofing module, before calculating the liveliness, the image quality is additionally estimated, while the quality threshold (default value is 30) allows excluding images with insufficient quality from the processing pipeline;
  • mask-estimator: the service detects if a person in the image is wearing a medical mask. The result is a True/False verdict with a numerical value of confidence;
  • quality-assessment-estimator: the service is designed to estimate the quality of a face image. The result is a list of detected faces with a detailed image quality analysis;
  • verify-matcher: the service compares two face images to determine if they belong to the same person;
  • face-detector-template-extractor: the service is designed to extract the biometric template from a face image.

Interaction with the services is performed via REST API. The following sections contain the detailed description of API requests and responses that provide the main functionality of Image API.