This self-hosted application, Wardrobe, addresses the need for intelligent wardrobe management by leveraging AI for automatic clothing tagging and personalized outfit recommendations. It targets users who prioritize data privacy and wish to integrate AI into their daily routines, offering a benefit of streamlined organization and daily style assistance without relying on cloud services.

How It Works

Wardrobe utilizes a photo-based approach where users upload images of their clothing. An AI vision model automatically extracts details such as color, pattern, and style. These tagged items are stored in a PostgreSQL database. For outfit recommendations, a backend service (FastAPI) integrates with various AI text models (via OpenAI-compatible APIs like Ollama or OpenAI) to suggest outfits based on extracted item data, current weather conditions (using Open-Meteo), and user-defined occasions or preferences. A background worker (arq) manages asynchronous AI job processing, ensuring a responsive user experience.

Quick Start & Requirements

• Primary install/run command: docker compose up -d followed by docker compose exec backend alembic upgrade head.
• Prerequisites: Docker and Docker Compose installed, at least 4GB RAM, an AI service (Ollama recommended for local use, or an OpenAI API key).
• Links: Ollama (https://ollama.ai), OpenAI API Keys (https://platform.openai.com/api-keys).
• Estimated setup time: Services take ~30 seconds to become healthy post-startup.

Highlighted Details

• AI-powered automatic clothing detail extraction and outfit generation.
• Flexible AI backend support: OpenAI, Ollama, LocalAI, or any OpenAI-compatible API.
• Features include scheduled notifications, family wardrobe management, wear tracking, and analytics.
• Optional background removal for item images using rembg or an HTTP provider.

Maintenance & Community

The project provides a CONTRIBUTING.md file for contribution guidelines and directs users to GitHub Issues and Discussions for support. Specific details regarding core maintainers, sponsorships, or dedicated community channels (like Discord/Slack) are not explicitly mentioned in the README.

Licensing & Compatibility

The project is licensed under the MIT License. This permissive license allows for commercial use and integration into closed-source projects without significant restrictions.

Limitations & Caveats

Setup requires familiarity with Docker and configuring AI services. While designed for self-hosting, users must manage their AI model dependencies and resource allocation. The effectiveness of AI features is dependent on the chosen AI model's capabilities and configuration.