Closed YuyaKobayashi closed 5 years ago
I tried to convert network from Keras.
The model file DenseASPP_pretraining.h5 is available in here.
DenseASPP_pretraining.h5
The graph image of the network outputed from Keras is network.jpg. The log of DMP tool is tool_output.txt.
network.jpg
tool_output.txt
The image shows that the network's output shape is [None, 64 128, 19] but the converted network's output shape is bigger.
output shape of Conv2D just after ZeroPadding2D seems to be wrong.
Conv2D
ZeroPadding2D
This is due to dilation is not parsed and not used in in/out size calculation.
dilation
I tried to convert network from Keras.
The model file
DenseASPP_pretraining.h5
is available in here.The graph image of the network outputed from Keras is
network.jpg
. The log of DMP tool istool_output.txt
.The image shows that the network's output shape is [None, 64 128, 19] but the converted network's output shape is bigger.