run_flask_service.py

# importing all required library
import flask
import io
from flask_cors import CORS
from PIL import Image
from ResNetPredict import ResNetPredict

app = flask.Flask(__name__)

# Cross Origin Resource Sharing (CORS), making the URL accessible from cross-origin AJAX
CORS(app)

'''
  Instantiate the defined class related to ResNet model, 
  you can add your own defined class here, for specific task.
'''
resNet_predict = ResNetPredict()

'''
  Loading the model on memory, this is efficient way to store your model 
  on memory along the way, rather than load on every endpoint call
'''
resNet_predict.load_model()

'''
  An endpoint for predicting an image using ResNet pre-trained model
  you can add more endpoint. Here you will define the endpoint of URL 
  and method what you will use
'''
@app.route("/predict_resnet", methods=["POST"])
def predict():
    data = {"success": False}
    if flask.request.method == "POST":
        # get paremeter named "image" as we only need one image file as parameter 
        if flask.request.files.get("image"): 
            # read the image in PIL format
            image = flask.request.files["image"].read()
            image = Image.open(io.BytesIO(image))
            image = resNet_predict.prepare_image(image, target=(224, 224))
            result_value = resNet_predict.predict(image)

            data["predictions"] = []

            for (imagenetID, label, prob) in result_value[0]:
                r = {"label": label, "probability": float(prob)}
                data["predictions"].append(r)
            data["success"] = True
            
    # return the result as JSON text
    return flask.jsonify(data)


if __name__ == "__main__":
    print("loading the payload...")
    print("starting server...")
    app.run(host="localhost")