Wardrobe AI

Wardrobe AI is a personal wardrobe management and outfit recommendation application built with React, Vite, SQLite, and locally-hosted AI models through Ollama.

The application allows users to:

• Build and manage a digital closet
• Add clothing manually or through AI-assisted image analysis
• Generate outfit recommendations based on available wardrobe items
• Track clothing combinations and outfit history
• Maintain complete ownership of wardrobe and AI data through local hosting

No cloud AI services or paid API subscriptions are required.

---

Features

Closet Management

• Add, edit, and remove clothing items

• Categorize items as:

  • Shirts / Tops
  • Pants
  • Shoes

• Store:

  • Name
  • Color
  • Pattern
  • Style / Cut
  • Color swatch
  • Image

AI-Powered Clothing Analysis

Upload an image and allow a locally-hosted vision model to:

• Identify clothing type
• Estimate color
• Detect patterns
• Suggest clothing attributes

The user can review and correct AI-generated information before saving.

Outfit Generator

Generate outfit combinations using:

• Current wardrobe inventory
• Color harmony principles
• Existing outfit history
• Clothing availability

Recommendations are generated entirely through a self-hosted AI model.

Local Data Ownership

Wardrobe data is stored locally in SQLite.

Benefits:

• No cloud storage required
• Full data ownership
• Fast performance
• Offline functionality on local network

---

Technology Stack

Frontend

• React
• Vite
• JavaScript

Backend

• Node.js
• Express

Database

• SQLite

AI

• Ollama
• llama3 (text generation)
• llava-llama3 (vision analysis)

Optional

• Docker
• Open WebUI

---

Folder Structure

wardrobe-ai/
│
├── src/
│   ├── api/
│   │   ├── aiClient.js
│   │   └── wardrobeApi.js
│   │
│   ├── App.jsx
│   └── main.jsx
│
├── server/
│   ├── server.cjs
│   ├── db.cjs
│   └── wardrobe.db
│
├── package.json
├── vite.config.js
└── README.md

---

Requirements

Software

Install:

• Node.js 22+
• Git
• Ollama
• Docker Desktop (optional)

---

Installation

1. Clone the Repository

git clone <repository-url>
cd wardrobe-ai

---

2. Install Frontend Dependencies

npm install

---

3. Install Backend Dependencies

npm install express cors sqlite3

---

4. Install Ollama

Download and install:

https://ollama.com

Verify installation:

ollama --version

---

5. Download AI Models

Text Model:

ollama pull llama3

Vision Model:

ollama pull llava-llama3

Verify:

ollama list

---

6. Start Ollama

ollama serve

---

7. Start Backend

From the project root:

npm run server

Expected output:

Wardrobe API running on port 3001
Connected to wardrobe database.

---

8. Start Frontend

Open a second terminal:

npm run dev

Expected output:

Local:
http://localhost:5173

---

Using on Other Devices

If devices are on the same Wi-Fi network:

Find your laptop IP:

ipconfig

Look for:

IPv4 Address

Example:

10.0.0.196

Open Wardrobe AI from another device:

http://10.0.0.196:5173

Examples:

• Tablet
• Android phone
• iPhone
• Secondary laptop

---

Data Storage

Wardrobe information is stored in:

server/wardrobe.db

This database contains:

• Clothing items
• Outfit history
• Future application data

Backing up this file preserves the user's wardrobe.

---

AI Configuration

Current default models:

Text:
llama3

Vision:
llava-llama3

Configuration can be modified in:

src/api/aiClient.js

---

Future Roadmap

Planned features:

• Outfit history analytics
• AI feedback learning system
• Shopping assistant mode
• Remote access outside home network
• User profiles
• Cloud backup options
• Mobile-first UI refinements
• Wardrobe statistics dashboard

---

License

This project is intended for personal learning and development purposes.

Modify and extend as needed.