Python API for Video Portrait Processing

I'm seeking a skilled Python developer to create a small API designed for video portrait extraction and natural-looking enhancement.

Main Tasks:

Input:
- Accept video upload (MP4 preferred format) via API.

Processing:
- Utilize MediaPipe (or similar) to:
- Detect the primary person in the video.
- Extract 3 high-quality frames.

- For each extracted frame:
- Generate:
- Thumbnails (small-sized images)
- Profile (larger-sized images )

- Apply subtle enhancement techniques to improve appearance while maintaining realism:
- Mild skin smoothing and wrinkle reduction (natural look only)
- Brightness and contrast adjustment
- Slight sharpening to improve clarity
- Softening of under-eye shadows
- Color correction for even skin tone and natural complexion
- Optional light teeth whitening (if visible)

Output:
- Return 3 thumbnails and 3 profile pictures (JPG or PNG).
- Include a JSON response containing:
- status: success/failure
- thumbnail_paths: [list of output image filenames]
- profile_paths: [list of output image filenames]
- flags: [e.g., 'no_face_detected', , 'low_resolution']

Ideal Skills and Experience:
- Proficiency in Python and experience with MediaPipe or similar tools for face detection.
- Strong understanding of video and image processing using OpenCV or PIL.
- Familiarity with basic enhancement techniques for portraits (e.g., smoothing, tone correction).
- Experience developing REST APIs (Flask or FastAPI preferred).
- Ability to deliver natural-looking improvements that subtly make subjects look ~5 years younger.

Let me know if you’d like a Docker container for easier deployment or example video/image files for testing.

Example video here : http://bit.ly/3GkkS3g

JSON Output Template
{
"status": "success",
"thumbnail_paths": ["thumb1.jpg", "thumb2.jpg", "thumb3.jpg"],
"profile_paths": ["profile1.jpg", "profile2.jpg", "profile3.jpg"],
"flags": []
}