The API was developed using Flask. It loads 3 models that I trained here using Monk library, receives pictures from the user and returns the predicted make, model and year.
- Clone this repository, or download as zip and extract to a directory
- Activate the virtual environment
- Install the requirements using
pip install -r requirements.txt
- Run the app using
python app.py
- To test the API using cURL
-
Install cURL using
pip install curl
-
Run the following command after updating the path to the image
curl -X POST -F file=@<image_path> 'http://0.0.0.0:5000/predict'
-
- To test the API using a user interface
- Open http://0.0.0.0:5000 in a web browser
- Upload an image using the webpage and press the predict button