virtual_camera_face_only.py

"""
BlurBerry Face-Only Virtual Camera Script

This script:
1. Runs camera with face enrollment
2. Starts virtual camera with face privacy protection only
3. Whitelists enrolled faces (your face stays clear)
4. Blurs all other faces
5. Streams to Zoom/Discord via OBS Virtual Camera
"""
import cv2
import numpy as np
import sys
import time
import os
from datetime import datetime

# Add project root to path
sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))

from shared_types import PipelineConfig
from blurberry.video.virtual_video_loop import VirtualVideoLoop
from blurberry.video.face_pipeline import FacePipeline

# Virtual camera import
try:
    import pyvirtualcam
    VIRTUAL_CAMERA_AVAILABLE = True
    print("✅ Virtual camera available")
except ImportError:
    VIRTUAL_CAMERA_AVAILABLE = False
    print("⚠️  Virtual camera not available, using local display")

FRAMES_NEEDED = 60   # collect 60 frames for enrollment

def face_only_virtual_camera():
    """Face-only virtual camera with privacy protection"""
    print("\n" + "="*60)
    print("🎭 BlurBerry AI - Face-Only Virtual Camera")
    print("="*60)
    print("\n🔒 Privacy Protection Features:")
    print("   ✅ Face Recognition (your face stays clear)")
    print("   ❌ Other Faces (auto-blurred)")
    print("\n📋 Controls:")
    print("   E - Enroll your face")
    print("   S - Skip enrollment, start virtual camera")
    print("   Q - Quit")
    print("\n🎥 Output: OBS Virtual Camera (for Zoom/Discord)")
    print("="*60)

    # Configuration with face detection only
    config = PipelineConfig(
        blur_faces=True,           # Enable face blur (whitelisted faces won't be blurred)
        blur_plates=False,         # Skip object detection
        blur_cards=False,          # Skip object detection
        blur_nsfw=False,           # Skip object detection
        blur_text_pii=False,       # Skip object detection
        detection_cadence=15,      # Not used for face-only
        blur_strength=51,          # Gaussian blur kernel size
        face_similarity_threshold=0.38  # Face recognition threshold
    )

    # Initialize face pipeline
    face_pipeline = FacePipeline(config)

    # Setup camera for enrollment
    cap = cv2.VideoCapture(0)
    cap.set(cv2.CAP_PROP_FRAME_WIDTH, 1280)
    cap.set(cv2.CAP_PROP_FRAME_HEIGHT, 720)

    collecting = False
    embeddings_collected = []
    enrolled = False

    # Enrollment phase
    while True:
        ret, frame = cap.read()
        if not ret:
            break

        display = frame.copy()
        faces = face_pipeline.face_app.get(frame)

        if collecting:
            if faces:
                face = max(faces, key=lambda f: (f.bbox[2]-f.bbox[0]) * (f.bbox[3]-f.bbox[1]))
                embeddings_collected.append(face.embedding)

                # Draw progress bar
                progress = len(embeddings_collected) / FRAMES_NEEDED
                bar_w = int(400 * progress)
                cv2.rectangle(display, (120, 680), (520, 700), (50, 50, 50), -1)
                cv2.rectangle(display, (120, 680), (120 + bar_w, 700), (0, 220, 100), -1)
                cv2.putText(display, f"Enrolling: {len(embeddings_collected)}/{FRAMES_NEEDED}",
                            (120, 675), cv2.FONT_HERSHEY_SIMPLEX, 0.6, (0, 220, 100), 2)

                # Draw box around face being enrolled
                x1, y1, x2, y2 = [int(v) for v in face.bbox]
                cv2.rectangle(display, (x1, y1), (x2, y2), (0, 220, 100), 2)
                cv2.putText(display, "Enrolling...", (x1, y1 - 8),
                            cv2.FONT_HERSHEY_SIMPLEX, 0.6, (0, 220, 100), 2)
            else:
                cv2.putText(display, "No face detected — look at camera",
                            (120, 450), cv2.FONT_HERSHEY_SIMPLEX, 0.6, (0, 100, 255), 2)

            # Auto-finish when enough frames collected
            if len(embeddings_collected) >= FRAMES_NEEDED:
                collecting = False
                face_pipeline.enroll_from_embeddings(embeddings_collected, name="owner")
                enrolled = True
                print(f"\n✅ [Enrollment] Done! Collected {len(embeddings_collected)} frames.")
                print("🎯 Your face is now whitelisted and will stay clear!")
        else:
            # Draw face boxes while idle
            for face in faces:
                x1, y1, x2, y2 = [int(v) for v in face.bbox]
                cv2.rectangle(display, (x1, y1), (x2, y2), (200, 200, 200), 1)

        # Status overlay
        status = ""
        if enrolled:
            status = "✅ Face enrolled! Press S to start virtual camera"
            color = (0, 220, 100)
        elif collecting:
            status = "📸 Collecting frames..."
            color = (0, 200, 255)
        else:
            status = "Press E to enroll your face | Press S to skip enrollment"
            color = (200, 200, 200)

        cv2.putText(display, status, (10, 30),
                    cv2.FONT_HERSHEY_SIMPLEX, 0.55, color, 2)

        # Privacy features status
        privacy_status = "🔒 Face Privacy: Your face clear, others blurred"
        cv2.putText(display, privacy_status, (10, 60),
                    cv2.FONT_HERSHEY_SIMPLEX, 0.5, (255, 165, 0), 2)

        # Virtual camera status
        if VIRTUAL_CAMERA_AVAILABLE:
            cv2.putText(display, "🎥 Virtual Camera Ready", (10, 90),
                        cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2)

        # Instructions
        cv2.putText(display, "E: Enroll | S: Start Camera | Q: Quit", (10, 720-20),
                    cv2.FONT_HERSHEY_SIMPLEX, 0.6, (255, 255, 255), 2)

        cv2.imshow("BlurBerry — Face-Only Virtual Camera", display)

        key = cv2.waitKey(1) & 0xFF

        if key == ord('q'):
            cap.release()
            cv2.destroyAllWindows()
            sys.exit(0)

        elif key == ord('e') and not collecting:
            embeddings_collected = []
            collecting = True
            print("\n📸 [Enrollment] Started collecting frames. Hold still and face the camera.")
            print("🎯 Your face will be whitelisted and stay clear in virtual camera!")

        elif key == ord('s'):
            cap.release()
            cv2.destroyAllWindows()
            break

    # ── Start Face-Only Virtual Camera Streaming ──
    print("\n" + "="*60)
    print("🚀 Starting Face-Only Virtual Camera...")
    print("="*60)
    
    if enrolled:
        print("✅ Your enrolled face will NOT be blurred.")
        print("🎯 All other faces WILL be blurred.")
    else:
        print("⚠️  No face enrolled. ALL detected faces will be blurred.")
    
    print("\n🔒 Face Privacy Protection Active:")
    print("   ✅ Your Face - Clear (if enrolled)")
    print("   ❌ Other Faces - Blurred")
    
    print("\n🎥 Virtual Camera: OBS Virtual Camera")
    print("📱 Compatible: Zoom, Discord, Teams, Meet")
    print("⏹️  Press Q to quit")
    print("="*60)

    # Create virtual video loop
    loop = VirtualVideoLoop(config, source=0, debug=False)

    # Pass the already-enrolled face pipeline into the loop
    loop.face_pipeline = face_pipeline

    # Note: No object detector added - face-only mode

    print("🎭 Face detection loaded:")
    print("   - Face recognition (your face clear)")
    print("   - Face blurring (other faces blurred)")

    try:
        # Start virtual camera stream with face-only privacy
        loop.run(use_virtual_camera=True)
    except KeyboardInterrupt:
        print("\n🛑 Stopped by user")
    except Exception as e:
        print(f"❌ Error: {e}")
        print("💡 Make sure OBS Studio is installed for virtual camera")
    finally:
        loop.stop()
        print("\n🔚 Virtual camera stream stopped")
        print("🎯 Face privacy protection disabled")
        print("="*60)

if __name__ == "__main__":
    print("🎭 BlurBerry AI - Face-Only Privacy Virtual Camera")
    print("="*60)
    
    # Check virtual camera availability
    if not VIRTUAL_CAMERA_AVAILABLE:
        print("⚠️  Warning: Virtual camera not available")
        print("   Install with: pip install pyvirtualcam")
        print("   Or install OBS Studio for virtual camera driver")
        response = input("Continue with local display? (y/n): ").strip().lower()
        if response != 'y':
            sys.exit(0)
    
    # Run face-only virtual camera
    face_only_virtual_camera()