TRINITY STYLE AI

A premium fashion e-commerce ecosystem featuring an integrated AI Virtual Try-on engine, bridging the gap between luxury retail and computer vision technology.

Overview

Trinity Style is a full-stack graduation project designed to redefine the online shopping experience. By combining a Luxury Minimalist aesthetic with state-of-the-art AI, the platform allows users to visualize garments on their own photos, reducing return rates and enhancing engagement.

✨ Core Features 🎨 Luxury Minimalist UI: High-contrast, black-and-white design language focused on premium brand identity

🤖 AI Virtual Try-on: Leverages SegFormer for precise segmentation and ControlNet to generate realistic clothing overlays

🔒 Secure Infrastructure: Features a robust PHP backend with OTP authentication and secure SQL transaction logic

📊 User Tier System: Specialized logic for loyalty tiers and personalized "Try-on History" archives

💻 Tech Stack

Frontend: HTML5, CSS3, JavaScript Backend: PHP, Python(Flask/FastAPI), MySQL

AI/ML: SegFormer, ControlNet, IP-Adapter Environment: macOS / Windows Cross-platform development

🛠 Installation & Setup

Clone the repository:

git clone https://github.com/trinity-newfed/Trinity-Style-AI.git

https://github.com/user-attachments/assets/907da0ec-a473-4ed2-826a-cde5ee686c04

Backend setup:

! This project uses Composer to manage PHP libraries. Make sure you have Composer installed before proceeding.

! For AI, Stable Diffusion and Qwen require venv and ollama

OLLAMA:

OLLAMA Download Link:

https://ollama.com/download/windows

MAILER

PHP Composer Download:

https://getcomposer.org/Composer-Setup.exe

Environment Configuration:

composer require vlucas/phpdotenv

Mailer Integration:

composer require phpmailer/phpmailer

REQUIRE FOR SEGFORMER

The AI engine requires Python 3.10.11. Note that the SegFormer model download exceeds 5.4GB on the first run. Python: https://www.python.org/downloads/release/python-31011/

🛰 Pipeline Architecture

1. SAM (Segment Anything): For initial segmentation.
2. Mask Generation: Creating the inpainting area.
3. ControlNet & IP-Adapter: To maintain garment texture and structure.
4. Stable Diffusion 1.5 Inpainting: Final image synthesis.

Virtual Environment Setup

cd AI python -m venv sd15

Activate environment

sd15\Scripts\activate

Activate on Windows (PowerShell)

If blocked, run:

1. Set-ExecutionPolicy -Scope Process -ExecutionPolicy Bypass
2. sd15\Scripts\Activate.ps1

Standard dependencies

pip install torch torchvision diffusers transformers safetensors accelerate

Specific CPU Torch Installation (Required for stability on non-GPU systems)

pip install torch==2.6.0 --index-url https://download.pytorch.org/whl/cpu

Database & API Bridge

pip install mysql-connector-python flask-cors

Requirements.txt

pip install -r requirements.txt

Documentation

🎬PHP MAILER SET UP VIDEO

1. Composer Set Up

https://github.com/user-attachments/assets/cc1d23ae-a700-4db6-ac55-361c17de06be

2. PHP Mailer Set Up

https://github.com/user-attachments/assets/bdad452c-ba02-48d5-ba3a-682ede1fb063

🎬PYTHON AND VIRTUAL ENVIRONMENT SET UP VIDEO

Python Download Set Up

https://github.com/user-attachments/assets/74b8635d-dc7a-4f48-8666-985813320484

Virtual Environment Set Up

https://github.com/user-attachments/assets/38f5672e-3a06-48e3-bdf1-04fc3ca82ca9

🎬OLLAMA SET UP VIDEO

https://github.com/user-attachments/assets/ebbee385-929e-48e3-ae1e-24bfc6bd5efc