https://github.com/bin-detective/bindetective-ml-backend
The Bindetective ML Service is a FastAPI-based backend for real-time waste image prediction in the Bin Detective app. It uses Robin, our custom ML model, to classify waste from image and is deployed on Google Cloud Run for serverless execution. The service integrates seamlessly with our main Express.js backend for enhanced and scalable performance
https://github.com/bin-detective/bindetective-ml-backend
fastapi python rest-api
Last synced: 2 months ago
JSON representation
The Bindetective ML Service is a FastAPI-based backend for real-time waste image prediction in the Bin Detective app. It uses Robin, our custom ML model, to classify waste from image and is deployed on Google Cloud Run for serverless execution. The service integrates seamlessly with our main Express.js backend for enhanced and scalable performance
- Host: GitHub
- URL: https://github.com/bin-detective/bindetective-ml-backend
- Owner: Bin-Detective
- Created: 2024-11-24T22:17:58.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2024-12-10T11:57:10.000Z (over 1 year ago)
- Last Synced: 2025-09-05T20:00:29.878Z (10 months ago)
- Topics: fastapi, python, rest-api
- Language: Python
- Homepage:
- Size: 63.5 KB
- Stars: 0
- Watchers: 0
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: readme.md
Awesome Lists containing this project
README
# Robin FastAPI Service
## Table of Contents
1. [Project Introduction](#project-introduction)
2. [Project Structure](#project-structure)
## Project Introduction
The Bindetective ML Service Backend is a dedicated service for machine learning-based waste image prediction, integrated into the Bin Detective app. Built using the FastAPI framework, this service is designed to handle real-time image predictions for waste classification. The service accepts image URLs, downloads the images, and utilizes a pre-trained machine learning model to classify the waste type.
The service is deployed on Google Cloud Run, ensuring scalable, serverless execution. It utilizes a custom model handler, WastePredictionServicer, which is responsible for running the prediction logic based on the input image. The app also includes robust error handling, logging, and a clean API interface to interact with the frontend of the Bin Detective app.
The backend is designed for efficiency, scalability, and security, providing high-performance predictions with minimal overhead. It seamlessly integrates with other services within the Bin Detective ecosystem, including the Express-based backend, ensuring a smooth flow of data and functionality across the platform.
## Project Structure
```plain
project_root/
│
├── src/
│ ├── utils/
│ │ └── model_utils.py
│ └── handler.py
│
├── .env
├── config.py Set static environment variables here
├── requirements.txt] Package dependencies, Use Python 3.10
├── Dockerfile
├── .gitignore
└── app.py
```