Child Face Prediction & Adoption Matching Service

This project aims to predict the face of a child based on parent photos and match adoption candidates with similar facial features. The ultimate goal is to encourage adoption in South Korea and create positive awareness about family and birth, addressing the low birth rate crisis.

📝 Project Summary

• Objective: Promote adoption and foster a positive perception of childbirth.
• Key Features:
  1. Generate an estimated child's face from parent photos.
  2. Match the generated face with available children for adoption.

By using facial prediction and matching technology, this project can help prospective parents envision their future children and promote adoption by providing personalized adoption recommendations.

---

🔧 Technologies Used

• Deep Learning: Utilizes the ResNet-18 model for image synthesis.
• Dataset: The family facial dataset is sourced from AIHub.
• Web Development: Frontend interface built using HTML, CSS, JavaScript.

---

🚀 How It Works

1. Parent Photo Upload:

   • Parents upload their photos through the web interface.

2. Child Face Prediction:

   • The system uses AI to predict and generate a potential child's face based on the parent photos.

3. Adoption Matching:

   • If a child with a similar appearance is available for adoption, the system notifies the parents via SMS.

---

🔍 Key Benefits

• Promote Adoption: Provides personalized adoption matching based on facial resemblance.
• Address Low Birth Rate: Supports national efforts to address the low birth rate crisis by promoting both adoption and childbirth.

---

📅 Future Plans

• Phase 1: AI model development and accuracy optimization.
• Phase 2: User interface (UI) design and mobile-friendly development.
• Phase 3: Data collection and model training.
• Phase 4: System testing and final integration.
• Phase 5: Beta launch and feedback collection.

---