Liveness Detection (PAD) - id3 Developer Portal

1. Prerequisites

id3 Face SDK (v4.0 or later)
Visual Studio 2019 or later
.NET 6.0 or later
Valid id3 Face SDK license with PAD module enabled

2. Setting Up the Project

Add the required NuGet package to your .NET project:

Install-Package Id3.Face.SDK

Add the following using directives to your code:

                    using Id3.Face.SDK.PAD;
                    using Id3.Face.SDK.Core;
                    using Id3.Face.SDK.Utils;
                

3. Initializing the PAD Processor

Create and configure the PAD processor:

                    // Initialize the SDK
                    var sdk = new FaceSDK();
                    sdk.Initialize("YOUR_LICENSE_KEY");
                    
                    // Create PAD processor configuration
                    var padConfig = new PADProcessorConfiguration
                    {
                        // Set the detection threshold (0.0 to 1.0)
                        Threshold = 0.5f,
                        
                        // Enable specific detection types
                        EnablePhotoAttackDetection = true,
                        EnableScreenAttackDetection = true,
                        EnableMaskAttackDetection = true,
                        
                        // Set maximum processing time (ms)
                        MaxProcessingTimeMs = 1000
                    };
                    
                    // Create the PAD processor
                    var padProcessor = new PADProcessor(padConfig);
                

Note: Replace YOUR_LICENSE_KEY with your actual license key.

4. Processing Frames

Process video frames to detect presentation attacks:

                    public PADResult ProcessFrame(byte[] frameData, ImageFormat format)
                    {
                        try
                        {
                            // Load the image
                            using (var image = ImageUtils.LoadImage(frameData, format))
                            {
                                // Process the frame
                                var result = padProcessor.Process(image);
                                
                                // Check if a face was detected
                                if (result.FaceDetected)
                                {
                                    // Check for presentation attack
                                    if (result.IsAttack)
                                    {
                                        Console.WriteLine($"Attack detected! Type: {result.AttackType}, " +
                                            $"Confidence: {result.Confidence:P2}");
                                    }
                                    else
                                    {
                                        Console.WriteLine("Live face detected");
                                    }
                                    
                                    // Get detailed scores
                                    Console.WriteLine($"Live score: {result.LiveScore:P2}");
                                    Console.WriteLine($"Attack score: {result.AttackScore:P2}");
                                }
                                
                                return result;
                            }
                        }
                        catch (Exception ex)
                        {
                            Console.WriteLine($"Error processing frame: {ex.Message}");
                            return new PADResult { Error = ex };
                        }
                    }
                

5. Handling the Results

Process the PAD results in your application:

                    // Example of handling PAD results
                    void HandlePADResult(PADResult result)
                    {
                        if (!result.FaceDetected)
                        {
                            Console.WriteLine("No face detected");
                            return;
                        }
                        
                        if (result.IsAttack)
                        {
                            switch (result.AttackType)
                            {
                                case AttackType.Photo:
                                    Console.WriteLine("Photo attack detected!");
                                    break;
                                case AttackType.Screen:
                                    Console.WriteLine("Screen attack detected!");
                                    break;
                                case AttackType.Mask:
                                    Console.WriteLine("Mask attack detected!");
                                    break;
                                default:
                                    Console.WriteLine("Unknown attack type detected!");
                                    break;
                            }
                            
                            // Handle the attack (e.g., show warning, log event, etc.)
                            HandleAttackDetected(result);
                        }
                        else
                        {
                            // Proceed with face recognition or other operations
                            Console.WriteLine("Live face verified");
                            ProcessVerifiedFace(result);
                        }
                    }
                

6. Best Practices

Important: For production use, always implement additional security measures and server-side validation.

Lighting: Ensure good, even lighting for accurate detection
Frame Rate: Maintain 15-30 FPS for optimal performance
Error Handling: Implement comprehensive error handling
User Guidance: Provide clear instructions to users during the process
Testing: Test with various devices and attack vectors

7. Troubleshooting

Common Issues:

No face detected: Check camera permissions and ensure the face is clearly visible
False positives/negatives: Adjust the threshold value based on your requirements
Performance issues: Reduce frame resolution or processing frequency

Need Help? Visit our community forums or contact support for assistance.