divyanshu3234/clothing-recommender ? reverse-engineered prompt

Build me a hybrid clothing recommendation app that takes a user photo and body measurements, then suggests clothing items that are a good visual and fit match.

I want the system to combine image similarity from deep learning embeddings with measurement based clustering, so it feels personalized instead of just generic. Please make it work as a simple backend I can run locally, with a few API endpoints to upload a user image, send measurements, and get recommended items back. If it helps, include a small sample dataset and a clear way to load the saved model files and embeddings.

Use the existing project structure if possible, and keep it easy to test from the notebook files too. If you need current library details or best practices for FastAPI, look them up online and use sensible defaults. Make sure the code is clean, documented, and ready to containerize later.