updated.py

resource_field = {"Message": fields.String,
                  "Image_Embedding": fields.List(fields.String),
                   "ID_Embedding": fields.List(fields.String),
                   "Confidence": fields.String,
                   "Round Trip Time": fields.String,
                   "Process First Image Time": fields.String,
                   "Process Second Image Time": fields.String,
                   "Generate and Compare Time": fields.String}



@ns.route('/imagepred')
class FaceNetPred(Resource):
    @ns.doc(description="Generating Embeddings for face using FaceNet",
     responses={200: "Success",
                400: "Bad Request",
                500: "Internal Server Error"
                })
    @ns.expect(image)
    def post(self):
        try:
            start_time = time.time()
            print("Start time is {}".format(start_time), flush=True)
            image_data = request.json
            pred_time = time.time()
            image_orig = image_data['image1']
            image_id = image_data['image2']

            #Decode Images Byte Array
            start_decoding_image = time.time()
            or_image = _read_image(image_orig)
            or_image = np.array(or_image)

            start_decoding_second_image = time.time()
            or_image = _read_image(image_orig)
            or_image = np.array(or_image)

            start_decoding_second_image = time.time()
            id_image = _read_image(image_id)
            id_image = np.array(id_image)


            #Process the image Embedding in to appraoch,
            # One on the original image using mtcnn face cropping and the other on the sent cropped
            #Image.
            #For the cropped image, the only thing that remains is prewhitenning
            end_decoding_images = time.time()
            message, img_emb, id_emb, confidence = id_face_pred(or_image, id_image)
            end_comparing_signatures = time.time()
            round_time_interval = "{}".format(time.time() - start_time)
            
            process_first_image_time = "{}".format(start_decoding_second_image - pred_time)

            process_second_image_time = "{}".format(end_decoding_images - start_decoding_second_image)

            gen_compare_signature_time = "{}".format(end_comparing_signatures - end_decoding_images)
            print("Time interval is {}".format(round_time_interval), flush=True)
            print("After Uploads time for prediction is {}".format(time.time() - pred_time), flush=True)
            data = {"Message":message, "Image_Embedding":list(img_emb), "ID_Embedding":list(id_emb),
                      "Confidence": confidence, "Round Trip Time": round_time_interval,
                      "Process First Image Time":process_first_image_time , "Process Second Image Time":process_seco$            
            response = marshal(data, resource_field)
            
            
            except Exception as inst:
                  return {"message": "Something wrong with incoming request."
                          "Original Message:; {}".format(inst)}, 400