kunle-xy / dive_into_dl_cv Goto Github PK

This model uses transfer learning architecture, VGG16, to identify and predict gender and age of a person in an image.

• pip install -r requirements.txt

click on the link to Google colab Link:

1. Ensure all dependencies are installed pip install fastapi uvicorn aiofiles jinja2

• uvicorn is a minimal low-level serve/application interface for setting up APIs
• aiofiles enables server to work asynchronously with requests

2. cd to the folder that contains script.py and server.py run on terminal: -uvicorn server:app--reload

3. To fetch predictions: run on terminal:

• curl -X POST "http://127.0.0.1:8000/predict" -H "accept: application/json" -H "Content-Type: multipart/form-data" -F "file=@/path to your image.png;type=image/png"

• Modern Computer Vision with Pytorch ( Ayyadevara & Reddy 2020)