flve-ts-sdk

TypeScript SDK for Face Liveness Verification Engine (FLVE)

Installation

npm install @labaid-ai/flve-sdk

Quick Start

import { createClient } from '@labaid-ai/flve-sdk';
import * as fs from 'fs';

// Create client pointing to HuggingFace Space
const client = createClient({
  baseUrl: 'https://farukhannan-flve.hf.space',
  headers: {
    // Add authentication token if required
    // 'Authorization': 'Bearer YOUR_HF_TOKEN'
  }
});

// Read image file
const imageBytes = fs.readFileSync('path/to/image.jpg');

// Enroll a face
const enrollResponse = await client.enroll({
  userId: 'user123',
  imageData: imageBytes,
  format: 'IMAGE_FORMAT_JPEG',
});

// Verify a face
const verifyResponse = await client.verify({
  referenceId: enrollResponse.referenceId,
  imageData: imageBytes,
  format: 'IMAGE_FORMAT_JPEG',
});

console.log('Match:', verifyResponse.isMatch);
console.log('Confidence:', verifyResponse.confidenceScore);

Note: This SDK connects to the FLVE HuggingFace Space at https://farukhannan-flve.hf.space. The space may take a few seconds to wake up from sleep mode on first request.

Features

• Face Enrollment: Register faces for verification
• Face Verification: Verify faces against enrolled references
• Liveness Detection: Check if face is live (anti-spoofing)
• eKYC: Electronic Know Your Customer with challenge-response
• Streaming: Real-time video verification support
• TypeScript: Full type safety and autocompletion

API Specifications

This SDK includes bundled API specifications in the specs/ directory:

• specs/openapi.yaml - REST API specification
• specs/proto/ - gRPC Protocol Buffer definitions

These are included for reference and documentation purposes.

API Reference

Client Methods

enroll(request: EnrollRequest): Promise<EnrollResponse>

Enroll a new face for verification.

verify(request: VerifyRequest): Promise<VerifyResponse>

Verify a face against an enrolled reference.

checkLiveness(request: LivenessRequest): Promise<LivenessResponse>

Check if a face is live (anti-spoofing).

startEKYC(request: StartEKYCRequest): Promise<StartEKYCResponse>

Start an eKYC session with challenges.

submitEKYCFrame(request: SubmitEKYCFrameRequest): Promise<SubmitEKYCFrameResponse>

Submit a frame during eKYC session.

completeEKYC(request: CompleteEKYCRequest): Promise<CompleteEKYCResponse>

Complete an eKYC session and get results.

getEKYCStatus(request: GetEKYCStatusRequest): Promise<GetEKYCStatusResponse>

Get the status of an eKYC session.

health(request?: HealthRequest): Promise<HealthResponse>

Check service health.

Configuration

Default Configuration

The SDK is pre-configured to use the HuggingFace Space:

• Base URL: https://farukhannan-flve.hf.space
• Timeout: 30000ms
• Retry: Enabled with 3 attempts

Custom Configuration

const client = createClient({
  baseUrl: 'https://farukhannan-flve.hf.space',
  timeout: 30000,
  retry: {
    enabled: true,
    maxAttempts: 3,
    backoffMs: 1000,
  },
  headers: {
    // Add HuggingFace authentication token if needed
    'Authorization': 'Bearer YOUR_HF_TOKEN',
  },
  interceptors: {
    request: (config) => {
      // Modify request before sending
      console.log('Request:', config);
      return config;
    },
    response: (response) => {
      // Process response
      return response;
    },
    error: (error) => {
      // Handle errors
      console.error('Request failed:', error);
    },
  },
});

Environment Variables

For local development or custom deployments, you can override the base URL:

const client = createClient({
  baseUrl: process.env.FLVE_API_URL || 'https://farukhannan-flve.hf.space',
});

License

MIT

Version

0.1.0

Generated by FLVE SDK Generator
This SDK is auto-generated and self-contained. All API specifications are bundled for portability.