Intel® Distribution of OpenVINO™ Toolkit
Community support and discussions about the Intel® Distribution of OpenVINO™ toolkit, OpenCV, and all things computer vision-related on Intel® platforms.
6108 Discussions

Conversion of tensorflow graph is successful but the result with movidius graph is strange


I have a tensorflow model at Input node is "input", output node is "output".


Here is the log


/usr/lib/python3/dist-packages/scipy/stats/ DeprecationWarning: Importing from numpy.testing.decorators is deprecated, import from numpy.testing instead.


from numpy.testing.decorators import setastest


/usr/local/bin/ncsdk/Controllers/Parsers/TensorFlowParser/ SyntaxWarning: assertion is always true, perhaps remove parentheses?


assert(False, "Layer type not supported by Convolution: " + obj.type)


mvNCCompile v02.00, Copyright @ Intel Corporation 2017


/usr/local/lib/python3.5/dist-packages/tensorflow/python/util/ DeprecationWarning: inspect.getargspec() is deprecated, use inspect.signature() instead shape: (1, 224, 224, 3) res.shape: (1, 1, 1, 11) TensorFlow output shape: (1, 1, 11) /usr/local/bin/ncsdk/Controllers/ UserWarning: You are using a large type. Consider reducing your data sizes for best performance Blob generated


My output node is a concatenated result of:


output1 = tf.reshape(endpoints['movidius_endpoints1'], shape=(1,1,1,9)) output2 = tf.reshape(endpoints['movidius_endpoints1'], shape=(1,1,1,2)) output = tf.concat([output1, output2], -1, name='concate') output = tf.identity(output, name='output')


However, when i run the inference on movidius, what i get is just a tensor of shape (2,)


Please help to take a look

0 Kudos
0 Replies