frame

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Sign In

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Issue - mnNCCompile for MaskRCNN network

Hi,

I'm having trouble successfully compiling my mask-rcnn model ...

compile command

mvNCCompile -s 12 X.pb -is 640 512 -in=input_image -on=mrcnn_mask/Reshape_1/shape -o graph

error out

[Error 5] Toolkit Error: Stage Details Not Supported: Dynamic input shape not supported

Verbose

root@emulcahy-Inspiron-5558:/ncsdk# mvNCCompile -s 12 X.pb -is 640 512 -in=input_image -on=mrcnn_mask/Reshape_1/shape -o graph
/usr/lib/python3/dist-packages/scipy/stats/morestats.py:16: DeprecationWarning: Importing from numpy.testing.decorators is deprecated, import from numpy.testing instead.
  from numpy.testing.decorators import setastest
/usr/lib/python3.5/importlib/_bootstrap.py:222: RuntimeWarning: numpy.dtype size changed, may indicate binary incompatibility. Expected 96, got 88
  return f(*args, **kwds)
mvNCCompile v02.00, Copyright @ Intel Corporation 2017
X.pb
2019-02-22 02:15:42.243562: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA
[Error 5] Toolkit Error: Stage Details Not Supported: Dynamic input shape not supported

I've been reading that RCNN is not supported?
Could i get some clarification on whats happening here?

Many thanks,
E

Comments

This discussion has been closed.