A multimodal fashion search and recommendation system using a fine-tuned ResNet50 for visual features and SentenceTransformers for text understanding.

• Python 3.9 or higher
• All dependencies installed in the virtual environment

The backend is powered by FastAPI and handles the search logic using the fine-tuned models. Run the following command in your terminal from the project root:

./venv/bin/python main.py

The server will start at http://localhost:8000.

Once the server is running, you can open the premium search interface:

1. Navigate to the ux folder.
2. Open index.html in any modern web browser (Chrome, Safari, Edge).
   • Alternatively, use the absolute path: /Users/dev_joshi/Desktop/project/fashion_recsys/ux/index.html

• Visual Search: Drag and drop or upload an image to find visually similar items.
• Natural Language Search: Search for products using descriptive terms like "red dress" or "striped shirt".
• Fine-tuned Precision: Uses a model specifically trained on fashion categories for superior accuracy.

• main.py: FastAPI server entry point.
• ux/: Modern frontend interface (HTML/CSS/JS).
• data/: Contains datasets and fine-tuned model weights.
• search.py: Core multimodal search logic.
• model.py: Model definitions and loading.