Image inference error

[NorthLatitudeOne]

Hi, can somebody help to have a look my problem, thanks!
Image inference in web menu test image inference:
Predict result: [{"error": "Inference error <class 'KeyError'>: 'input_image'"}, 400]

Image inference in command line:
`E:\Anaconda\Lib\site-packages\simple_tensorflow_serving>curl -X POST -F 'image=images/NG0093.jpg' -F "model_version=3" 127.0.0.1:8500

<title>500 Internal Server Error</title>

Internal Server Error

The server encountered an internal error and was unable to complete your request. Either the server is overloaded or there is an error in the application.

`

Simple tensorflow serving log:
2019-04-18 09:00:46 ERROR Need to set image or images for form-data 2019-04-18 09:00:46 ERROR Exception on / [POST] Traceback (most recent call last): File "e:\anaconda\lib\site-packages\flask\app.py", line 2292, in wsgi_app response = self.full_dispatch_request() File "e:\anaconda\lib\site-packages\flask\app.py", line 1815, in full_dispatch_request rv = self.handle_user_exception(e) File "e:\anaconda\lib\site-packages\flask_cors\extension.py", line 161, in wrapped_function return cors_after_request(app.make_response(f(*args, **kwargs))) File "e:\anaconda\lib\site-packages\flask\app.py", line 1718, in handle_user_exception reraise(exc_type, exc_value, tb) File "e:\anaconda\lib\site-packages\flask\_compat.py", line 35, in reraise raise value File "e:\anaconda\lib\site-packages\flask\app.py", line 1813, in full_dispatch_request rv = self.dispatch_request() File "e:\anaconda\lib\site-packages\flask\app.py", line 1799, in dispatch_request return self.view_functions[rule.endpoint](**req.view_args) File "e:\anaconda\lib\site-packages\simple_tensorflow_serving\server.py", line 184, in decorated return f(*decorator_args, **decorator_kwargs) File "e:\anaconda\lib\site-packages\simple_tensorflow_serving\server.py", line 308, in inference json_result, status_code = do_inference() File "e:\anaconda\lib\site-packages\simple_tensorflow_serving\server.py", line 349, in do_inference if "model_name" in json_data: TypeError: argument of type 'NoneType' is not iterable 2019-04-18 09:00:46 INFO 127.0.0.1 - - [18/Apr/2019 09:00:46] "�[1m�[35mPOST / HTTP/1.1�[0m" 500 -

[jeffin07]

@NorthLatitudeOne i got a similar error
'error': "Inference error <class 'KeyError'>: 'in'"}
Did you find any solution for this ?
I also tried with the given model in Readme still the same type of error
{"error": "Inference error <class 'KeyError'>: 'features'"}

[serlina]

I used docker to run the simple_tensorflow_serving in local, and I can open the 127.0.0.1:8500
docker run -d -p 8500:8500 tobegit3hub/simple_tensorflow_serving
the python version is python2.7

while I used postman to send the post request for image inference client call, it also raise exception as below:

while the code for converting jpeg to base64 is :
from PIL import Image
import cv2
import cStringIO
import base64

def base64_encode_img(img):
"""
:param img:
:return:
"""
img_rgb = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
print('shape of img_rgb:', img_rgb.shape)
pil_img = Image.fromarray(img_rgb)

buf = cStringIO.StringIO()
pil_img.save(buf, format="JPEG", quality=100)
b64code = base64.urlsafe_b64encode(buf.getvalue()) #web-safe
# b64code = base64.b64encode('abcdefgdisoaufd,0.342,0.456,0.987')  # not web-safe
print(b64code)
return b64code

if name == "main":
img_BGR = cv2.imread('./mew.jpg')
base64_encode_img(img_BGR)

could someone give me help about the image client call for the model

[serlina]

sorry ,correct my above code, I just used below code to convert local jpg to base64, and then used the base64 in postman (post body)

with open("./mew.jpg", "rb") as image_file:
encoded_string = base64.b64encode(image_file.read())
print encoded_string

comments: if I use base64.urlsafe_b64encode to encode, it will raise another error

[tobegit3hub]

Hi @serlina , it depends on your TensorFlow SavedModel's op. If you use tf.decode_base64(model_base64_placeholder) to process the input data, you may try this client code which has test in our environment.

import requests
import base64

def main():
  image_file_name = "../../images/mew.jpg"
  image_b64_string = base64.urlsafe_b64encode(
      open(image_file_name, "rb").read())

  endpoint = "http://127.0.0.1:8500"
  input_data = {
      "data": {
          "image": [image_b64_string]
      }
  }
  result = requests.post(endpoint, json=input_data)