merge_lips.py

import argparse

from pipeline import merge_lips


def main() -> None:
    parser = argparse.ArgumentParser(description="Step 4: Merge lip-synced face region back into original video.")
    parser.add_argument("--original-video", required=True, help="Path to original video")
    parser.add_argument("--lip-synced-video", required=True, help="Path to lip-synced face video from external module")
    parser.add_argument("--face-coords", required=True, help="Path to face coordinates pkl")
    parser.add_argument("--lip-coords", required=True, help="Path to lip coordinates pkl")
    parser.add_argument("--output", default="final_synced_output.mp4", help="Output final merged video path")
    parser.add_argument("--audio", default=None, help="Optional external audio path from Step 3")
    args = parser.parse_args()

    final_path, audio_used = merge_lips(
        original_video_path=args.original_video,
        lip_synced_video_path=args.lip_synced_video,
        face_coords_path=args.face_coords,
        lip_coords_path=args.lip_coords,
        final_output_path=args.output,
        audio_path=args.audio,
    )

    print(f"Final merged video: {final_path}")
    print(f"Audio source used:  {audio_used}")


if __name__ == "__main__":
    main()