Skip to content

Lens

  • Purpose: Guide to quickly start using the Lens API.
  • Description: This API service provides advanced facial analysis by comparing images accessible via URLs. By analyzing the faces detected in each image, the service identifies and evaluates key facial features, expressions, and similarities or differences between the images. This can be particularly useful for applications in security, authentication, and image recognition.

Prerequisites

  • Is necessary to already have an API Credentials from your organization
  • Your account should have at least the roles: lens and usage.

Lens flow

Basic Face verification flow:

  1. The client sends a POST request to the Lens API at the /lens/face-verification endpoint.
  2. The Lens API creates a log USAGE record of the request.
  3. The Lens API extracts faces from the provided images.
  4. The Lens API applies face verification procedures to the extracted faces.
  5. The Lens API calculates face similarities based on the verification.
  6. The Lens API sends the analysis results back to the client.

Extracto flow

Usage

Files formats allowedFiles Maximum size
Images format: PNG, WEBP, JPEG/JPG100 MB

Limitations

  • PDF documents are not supported, this feature will be included in the future.
  • URL path with security are not allowed.
  • Disclaimer: Validation of identical twins is a complex task. Our AI model compares the similarities between two faces, and since faces are not as unique as fingerprints, the probability of high similarity between twins is significant. In the following Apple Forum Discussion, it is analyzed that even advanced tools using different sensors, such as the LiDAR sensor in iPhone FaceID, find it challenging to distinguish between identical twins.

Resources