Skip to content

PowerMind VisionID Overview

Purpose

PowerMind VisionID is a face recognition application for enrolling people, identifying them from photos or live camera capture, verifying a claimed identity, and managing authenticated user access to those workflows.

The application combines three concerns:

  • biometric enrollment and matching
  • browser-based live camera workflows with liveness checks
  • administrative access control for users and audit events

Business Use Cases

VisionID supports the following practical use cases:

  • employee or member enrollment using a single face photo
  • identity confirmation during check-in or access control
  • live attendance or visitor verification from a webcam or phone camera
  • identity verification against a known person ID
  • controlled delegation of operator access through supported user and administrator roles
  • audit logging of administrative actions

Key Features

  • authenticated access and role-based permissions
  • browser UI for sign-in, enrollment, and live identification
  • single-photo enrollment from JPG or PNG uploads
  • photo-based identification and verification
  • live-video identification with liveness detection
  • delete enrollment support
  • user and enrollment administration
  • audit history for security-sensitive actions
  • API access for authorized integrations

Supported Workflows

Authentication

  1. A user signs in with username and password.
  2. The API returns a signed access token with role information.
  3. The token is supplied in the Authorization header for subsequent requests.

User Administration

  1. A privileged user lists current accounts.
  2. A privileged user creates or deletes lower-privilege accounts.
  3. Password changes are controlled by role rules and current-password checks for self-service changes.

Face Enrollment

  1. An authorized operator uploads a photo and a unique person_id.
  2. The server detects exactly one face.
  3. The server stores the resulting embedding and enrollment metadata.

Face Identification

  1. An authorized user uploads a face photo or live frame sequence.
  2. The server runs liveness checks for live-video requests.
  3. The server compares embeddings against enrolled people.
  4. The server returns the best match and confidence score.

Face Verification

  1. An authorized user supplies a known person_id and a face photo.
  2. The server checks whether the person is enrolled.
  3. The server returns a matched or unmatched verdict for that identity.

High-Level Architecture

At a high level, VisionID has four functional layers:

  1. Browser UI or external API client
  2. Authentication and role-based access controls
  3. Protected identity and audit storage
  4. Face-recognition and liveness processing

PowerMind AI Studio integration

VisionID is available as a tool in PowerMind AI Studio. In Studio, operators open Tools → VisionID and use the Dashboard, Identities, Enrollment, Recognition, Verification, and Audit workspaces. Studio supplies the authenticated shell, project context, service-health indicator, and navigation; VisionID supplies the biometric workflows and results. See the User Guide for the screen-by-screen workflow.