Classification-REST-API

simple Keras imagenet classification Flask REST API , VGG , ResNet50 , InceptionV3 , Xception .

• to run : assume tenserflow already installed
  requirements :

$ pip install keras
$ pip install numpy
$ pip install Pillow
$ pip install flask gevent requests

• Starting the Keras server clone this repo and put you image within server file

start The Flask + Keras server , access the REST API via http://127.0.0.1:5000 :

$ python server.py

Requests can be submitted via curl or by running simple_request.py , case of using curl response will be json object :

$ curl -X POST -F image=@image.jpg 'http://localhost:5000/predict?model=VGG16'

specify your image path after image=@ and the name of your desired classifier after ?model= , by running simple_request.py ResNet50 is the defaul model