AI-First Album, Chat with your gallery album using plain language!
We update frequently, feel free to star us.
Introduction
Welcome to Album AI, an innovative project that takes your gallery management to the next level! Leveraging the power of the gpt-4o-mini and Haiku visual model, Album AI automatically detects and generates metadata for images in your album. Imagine being able to have a conversation with your photos; that’s what this project is all about!
Story
As a photography lover swamped with terabytes of photos, managing everything seemed like climbing a mountain. Traditional photo management tools were cumbersome, requiring too much time and effort. That’s when I discovered the potential of Haiku and gpt-4o-mini. With my partner, we created the first version in less than 24 hours, and the result is Album AI. We believe you’ll find it useful too!
Live Demo
Experience Album AI firsthand with our live demo!
Features
- Automatically discover images in your album utilizing a PgSQL database for storage.
- Use GPT-4-o-mini to automatically generate images’ metadata.
- Employ OpenAI’s Embedding API for metadata vectorization.
- Two main APIs:
- Search API: Allows you to type a query and fetch the most relevant images.
- Chat API: Facilitates a RAG conversation, retrieving images and generating responses.
- One-click deployment supports Docker container deployment on platforms like Render.
- Open-source license for integration and modification (contact us for commercial use).
How to Start Using Album AI?
For best results, it’s recommended to run Album AI locally. If you prefer to run it on a server, please deploy it yourself. Here’s how to get started:
1. Clone the Project
bash
git clone git@github.com:gcui-art/album-ai.git
cd album-ai
2. Modify the .env File
bash
cp .env.prod.example .env.prod
Open .env.prod with your favorite editor, then modify the configuration:
HOST_NAME= # Your local IP address, usually 192.168.x.x:8080
PROXY_URL= # (Optional) Your local proxy IP address, usually 192.168.x.x:7890, needed when accessing OpenAI API directly is not available
OPENAI_API_KEY= # Your OpenAI API key
ANTHROPIC_API_KEY= # Your Anthropic API key
3. Build and Run the Project
bash
chmod a+x .build.sh
.build.sh
4. Enjoy!
Open your browser and visit http://localhost:8080 to see the demo!
5. Add New Photos
Simply add new photos to the ‘images’ directory within the project, and the system will automatically recognize and vectorize the metadata. Afterwards, you can interact with your pictures using both search and chat functionalities.
API Reference
Album AI currently implements the following APIs:
GET /api/v1/file/search: Searches for images.POST /api/v1/chat: Chat with images.
Contribution
If you’d like to support Album AI, here are a few ways to do so:
- Fork the project and submit a Pull Request (PR).
- Submit an issue for any suggestions or bug reports.
- Spread the word by recommending the project to others and star it on GitHub.
Troubleshooting
If you encounter any issues or have questions, please submit them on our Github Issues page. We appreciate any feedback and will address it as soon as possible.
For more insights, updates, or to collaborate on AI development projects, stay connected with fxis.ai.
Disclaimer
If you intend to use Album AI for commercial purposes, please reach out to us.
Final Thoughts
At fxis.ai, we believe that such advancements are crucial for the future of AI, as they enable more comprehensive and effective solutions. Our team is continually exploring new methodologies to push the envelope in artificial intelligence, ensuring that our clients benefit from the latest technological innovations.