PyTorch Implementation: Age-Controlled Facial Editing

I need an experienced developer to build a prototype for controllable facial age editing using ready-made pretrained models only (no model training). The design is already defined; I need help with implementation.

What the system should do:
- Take a single face image and generate the same person at different target ages (e.g., 40, 60, 80) and at a younger age.
- - Apply optional local edits to specific facial regions (forehead, glabella, cheeks, perioral area, lips) using region masks, e.g., adding wrinkles or pigmentation, smoothing wrinkles, or increasing volume. Each region edit must be switchable on/off independently, and the system must also support a "no edit" mode (age change only).
- Preserve the person's identity across all outputs.

Technical scope (pretrained components only):
- Stable Diffusion XL with InstantID or IP-Adapter FaceID for identity-preserving age editing
- SDXL Inpainting for region-level edits
- Pretrained face parsing (e.g., BiSeNet) and facial landmarks to generate region masks
- Wrinkle and pigmentation measurement per region (before/after) using existing pretrained tools or simple image analysis
- Identity similarity (ArcFace) and age estimation scores for each output

Deliverables:
- Clean, documented Python code (GitHub repository)
- One inference script: input image + target age + selected region edits -> output images + scores
- Short README explaining setup and usage

Required skills: Python, PyTorch, Hugging Face Diffusers, Stable Diffusion XL, InstantID/IP-Adapter, inpainting, face parsing.

Please include in your bid:
- Links to previous Stable Diffusion / InstantID / inpainting projects
- Estimated delivery time

All edit settings (regions, prompts, edit strength) must be configurable through a config file, so I can adjust them myself after delivery.