extract_faces.py

import os
from PIL import Image
from retinaface import RetinaFace
from concurrent.futures import ProcessPoolExecutor

Image.MAX_IMAGE_PIXELS = 200000000  # Allow images up to 200 million pixels

DATA = [
    ((512, 512), r"G:\man dataset\man_4502_imgs_raw"),
    ((1024, 1024), r"G:\man dataset\man_4502_imgs_raw")
]

NUM_THREADS = 1  # Modify this for the desired number of threads
OUTPUT_ROOT = r"H:\man dataset\resize_faces"  # Root folder to save resized face images

## change resize enabled to resize
def resize_face(args, resize_enabled=False):
    try:
        input_path, resolution, output_root = args
    except ValueError as e:
        print(f"Error unpacking arguments: {e}")
        return

    try:
        image = Image.open(input_path)
    except Exception as e:
        print(f"Error loading image: {e}")
        return

    try:
        # Detect faces
        faces = RetinaFace.detect_faces(input_path)

        if not faces:
            print(f"No face detected in {input_path}")
            return

        # Assuming we take the first detected face
        face = next(iter(faces.values()))
        x, y, x2, y2 = face['facial_area']
        face_image = image.crop((x, y, x2, y2))

        if resize_enabled:
            # Resize the cropped face only if resizing is enabled
            face_image = face_image.resize(resolution, Image.LANCZOS)
    except Exception as e:
        print(f"Error in face detection or resizing: {e}")
        return

    try:
        output_folder = os.path.join(output_root, f"{resolution[0]}x{resolution[1]}" if resize_enabled else "no_resize")
        if not os.path.exists(output_folder):
            os.makedirs(output_folder)

        output_path = os.path.join(output_folder, os.path.basename(input_path))
        face_image.save(output_path, quality=100)
    except Exception as e:
        print(f"Error saving image: {e}")

def process_folder(input_folder, resolution, resize_enabled=False):
    image_paths = [os.path.join(input_folder, fname) for fname in os.listdir(input_folder) if fname.lower().endswith(('png', 'jpg', 'jpeg'))]
    
    total_images = len(image_paths)
    processed_count = 0

    with ProcessPoolExecutor(max_workers=NUM_THREADS) as executor:
        for _ in executor.map(resize_face, [(path, resolution, OUTPUT_ROOT) for path in image_paths], [resize_enabled] * len(image_paths)):
            processed_count += 1
            print(f"Processed {processed_count}/{total_images} images for resolution {resolution[0]}x{resolution[1]} (Resize: {'Enabled' if resize_enabled else 'Disabled'})...")

if __name__ == "__main__":
    resize_enabled = False  # Change this to True if resizing is needed
    for resolution, folder in DATA:
        process_folder(folder, resolution, resize_enabled)
    print("Processing complete!")