While accessing the output of mask detection model i am not getting the result.

Kamma__Vijayakumar · ‎05-07-2020

I have trained the model with IMAGEAI library with that i have got .h5 file anf .json file.

then to convert into openvino models i have done the below steps.

1)H5 to Tensorflow frozen

2)Frozen to Bin,xml by using mo_tf.py

got success till this part.

now while executing the model with exec_net = load_to_IE(model) the result showing NDArray , i am not understanding how to access my output.

{'conv_81/BiasAdd/Add': array([[[[-1.66312973e+02, -2.46958740e+02, -1.74602142e+02, ...,
-2.83901031e+02, -2.21121582e+02, -9.00830765e+01],
[-1.20000076e+02, -2.66399963e+02, -1.58648590e+02, ...,
-2.18675812e+02, -2.38010468e+02, -1.37554138e+02],
[-1.06553749e+02, -2.97883667e+02, -1.27020409e+02, ...,
-1.77826111e+02, -2.24372299e+02, -1.44611572e+02],
...,
[-4.78074760e+01, -1.75183441e+02, -1.38157776e+02, ...,
-4.33831978e+01, -4.24874916e+01, -6.06762924e+01],
[-1.84533119e+01, -1.68425415e+02, -1.11759216e+02, ...,
-6.22600098e+01, -5.67313538e+01, -4.46702423e+01],
[ 2.93867760e+01, -6.70677872e+01, -5.70977402e+01, ...,
-4.19528961e+01, -3.55563469e+01, -2.01602802e+01]],

[[ 1.40633591e+02, 1.15974365e+02, 1.17529114e+02, ...,
1.81284317e+02, 1.42111740e+02, 7.96556702e+01],
[ 1.27407059e+02, 2.11995117e+02, 1.85863602e+02, ...,
2.60103180e+02, 2.27830368e+02, 1.25303055e+02],
[ 1.64037766e+02, 1.96629532e+02, 1.11918442e+02, ...,
1.56511246e+02, 1.54128342e+02, 8.48983231e+01],
...,
[ 6.33352051e+01, 6.37791214e+01, 7.07268333e+00, ...,
-4.78412857e+01, -4.94226074e+01, -4.03650360e+01],
[-2.00031033e+01, 6.08664474e+01, 8.17405510e+00, ...,
-4.89005280e+01, -5.90939636e+01, -4.14923592e+01],
[-1.53922958e+02, -4.38242569e+01, -4.53684692e+01, ...,
-7.21609192e+01, -6.14896317e+01, -3.65516281e+01]],

[[-4.73735008e+01, -3.41962051e+01, -3.74608269e+01, ...,
-7.45596504e+00, -7.31755733e+00, -1.60885124e+01],
[-6.24234428e+01, -2.96515732e+01, -4.65041237e+01, ...,
-4.72871256e+00, -2.03644466e+00, -9.75383663e+00],
[-7.78091431e+01, -6.62882233e+01, -8.26859665e+01, ...,
-7.36377258e+01, -4.45033646e+01, -4.27225037e+01],
...,
[-5.21416435e+01, -4.79345245e+01, -1.94807243e+01, ...,
-4.63317719e+01, -5.14390182e+01, -4.86955376e+01],
[-5.49678001e+01, -4.29327850e+01, -1.45485783e+01, ...,
-3.20624237e+01, -2.77655869e+01, -3.03304882e+01],
[-3.63332634e+01, -5.35129395e+01, -2.23160210e+01, ...,
-3.47856407e+01, -3.04427185e+01, -2.80112000e+01]],

...,

[[-4.74601059e+01, -7.61016312e+01, -6.27142792e+01, ...,
-5.15001526e+01, -1.97596245e+01, 1.56756973e+00],
[-1.09927811e+02, -1.40803879e+02, -1.29083405e+02, ...,
-9.53482056e+01, -1.30481453e+01, 1.54791842e+01],
[-1.52028290e+02, -2.05058655e+02, -1.97106476e+02, ...,
-1.51681152e+02, -3.56237297e+01, 1.12130127e+01],
...,
[-8.74849091e+01, -4.64156952e+01, -1.48516827e+01, ...,
-5.56459236e+01, -1.25644817e+01, 1.05311594e+01],
[-8.36514206e+01, -5.47760010e+01, -1.75369740e+01, ...,
-1.92452297e+01, 6.29400921e+00, 1.16010056e+01],
[-8.55055695e+01, -6.86285858e+01, -3.49647751e+01, ...,
-2.27549591e+01, -8.11953735e+00, 2.38107443e-01]],

[[-7.85665161e+02, -8.74612976e+02, -9.88772461e+02, ...,
-1.27541602e+03, -9.81773193e+02, -5.76246216e+02],
[-1.26309900e+03, -1.26470850e+03, -1.29349731e+03, ...,
-1.47686731e+03, -1.09906421e+03, -6.09180664e+02],
[-1.78284009e+03, -1.69450464e+03, -1.49403320e+03, ...,
-1.78028442e+03, -1.26287354e+03, -7.03716125e+02],
...,
[-1.00871576e+03, -8.21046570e+02, -6.36620422e+02, ...,
-9.28006348e+02, -7.18372314e+02, -4.14899475e+02],
[-9.60229553e+02, -8.18582581e+02, -5.88664917e+02, ...,
-6.90466431e+02, -5.43034790e+02, -3.17665497e+02],
[-8.69738403e+02, -8.36209045e+02, -5.64757874e+02, ...,
-5.97556946e+02, -4.59418213e+02, -2.51658508e+02]],

[[-4.45263901e+01, -7.05552368e+01, -1.44238162e+00, ...,
-3.45628738e+01, -5.43552780e+01, -4.13258972e+01],
[-8.26740570e+01, -1.21586502e+02, -1.14394188e+01, ...,
-5.72901840e+01, -7.67189178e+01, -5.92793922e+01],
[-1.42641937e+02, -1.68532562e+02, -4.49021759e+01, ...,
-9.04630585e+01, -7.94408722e+01, -5.09029770e+01],
...,
[-9.37536774e+01, -6.21579819e+01, 3.14342213e+01, ...,
3.28904057e+00, -1.52875957e+01, -4.29225874e+00],
[-7.81461029e+01, -7.22118607e+01, 1.63341351e+01, ...,
7.96173859e+00, 2.67618060e+00, 5.32587767e+00],
[-6.89589844e+01, -4.00298271e+01, -2.01719403e-01, ...,
-5.96850276e-01, 8.10801268e-01, 3.34924364e+00]]]],
dtype=float32)}

please help me on this.

Kamma__Vijayakumar · ‎05-07-2020

my first model i have trained my model with custom data with imageai library on pre_trained_yolo.h5 file.

Kamma__Vijayakumar · ‎05-07-2020

i have checked my openvino xml file for checking the output layers. i dont see any proper output layers got created. Please help me asap.

Munesh_Intel · ‎06-04-2020

Hi Vijayakumar,

Thank you for reaching out.

It’s unclear as to why you were unable to properly convert the trained model to Intermediate Representation (IR), without many details of the model and command used to convert.

Please share more information about your model, the topology that it is based on, the layers used if it's a custom model, command given to Model Optimizer to convert the trained model to IR, and also environment details (versions of OpenVINO, OS, Python, TF, CMake, etc.).

The following page, ‘Preparing and Optimizing Your Trained Model’, contains relevant information that will be helpful to you.

https://docs.openvinotoolkit.org/2020.3/_docs_MO_DG_prepare_model_Prepare_Trained_Model.html

Regards,

Munesh