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!

Alert: This site has migrating to our new home at https://forums.intel.com/s/topic/0TO0P000000PqZDWA0/intel-neural-compute-sticks. Please visit us there!

mvNCProfile fails with Error 25 - 5x5 Convolution outputChannels Dimension too small

Hi,
I am trying to run a caffe model for anti-aliasing (AA) for inclusion in rendering pipeline - replicating work from Deep Shading paper. The idea is to compare performance of trained model with traditional methods of AA, or other graphics techniques.

The sample prototxt is below - 2 conv and 2 RELU layers. Using mvNCCompile on this works fine, but mvNCProfile fails with [Error 25] Myriad Error: "5x5 Convolution outputChannels Dimension too small.".

Any help in solving this is appreciated.

aaforNC.prototxt

name: "ShadingNet"

input: "data"
input_shape {
dim: 1
dim: 6
dim: 256
dim: 256
}

layer {
name: "down_level_0_conv_top"
type: "Convolution"
bottom: "data"
top: "down_level_0_conv_top"
#blobs_lr: 1
#blobs_lr: 2
convolution_param {
num_output: 8
kernel_size: 5
pad: 2
stride: 1
engine: CAFFE
weight_filler {
type: "gaussian"
std: 0.011
}
bias_filler {
type: "constant"
value: 0
}
}
}

layer {
name: "down_level_0_relu"
type: "ReLU"
bottom: "down_level_0_conv_top"
top: "down_level_0_relu"
relu_param{
negative_slope: 0.01
engine: CAFFE
}
}

layer {
name: "up_level_0_conv_top"
type: "Convolution"
bottom: "down_level_0_relu"
top: "up_level_0_conv_top"
#blobs_lr: 1
#blobs_lr: 2
convolution_param {
num_output: 1
kernel_size: 5
pad: 2
stride: 1
engine: CAFFE
weight_filler {
type: "gaussian"
std: 0.01
}
bias_filler {
type: "constant"
value: 0
}
}
}

layer {
name: "up_level_0_relu"
type: "ReLU"
bottom: "up_level_0_conv_top"
top: "result"
relu_param{
engine: CAFFE
negative_slope: 0.01
}
}

layer {
name: "loss"
type: "Softmax"
bottom: "result"
top: "loss"
}

Message and Error Log
spl@spl-Blade:~/Desktop$ mvNCCompile netMV.prototxt
mvNCCompile v02.00, Copyright @ Movidius Ltd 2016

/usr/local/lib/python3.5/dist-packages/matplotlib/backends/backend_gtk3.py:58: DeprecationWarning: Gdk.Cursor.new is deprecated
cursors.MOVE : Gdk.Cursor.new(Gdk.CursorType.FLEUR),
/usr/local/lib/python3.5/dist-packages/matplotlib/backends/backend_gtk3agg.py:18: UserWarning: The Gtk3Agg backend is known to not work on Python 3.x with pycairo. Try installing cairocffi.
"The Gtk3Agg backend is known to not work on Python 3.x with pycairo. "
/usr/local/lib/python3.5/dist-packages/scipy/lib/decorator.py:219: DeprecationWarning: inspect.getargspec() is deprecated, use inspect.signature() instead
first = inspect.getargspec(caller)[0][0] # first arg
/usr/local/lib/python3.5/dist-packages/scipy/optimize/nonlin.py:1498: DeprecationWarning: inspect.getargspec() is deprecated, use inspect.signature() instead
args, varargs, varkw, defaults = inspect.getargspec(jac.init)
/usr/local/lib/python3.5/dist-packages/scipy/stats/_distn_infrastructure.py:611: DeprecationWarning: inspect.getargspec() is deprecated, use inspect.signature() instead
sign = inspect.getargspec(self._stats)
/usr/local/lib/python3.5/dist-packages/scipy/stats/_distn_infrastructure.py:648: DeprecationWarning: inspect.getargspec() is deprecated, use inspect.signature() instead
shapes_args = inspect.getargspec(meth)
****** WARNING: using empty weights ******
spl@spl-Blade:~/Desktop$ mvNCProfile netMV.prototxt
mvNCProfile v02.00, Copyright @ Movidius Ltd 2016

/usr/local/lib/python3.5/dist-packages/matplotlib/backends/backend_gtk3.py:58: DeprecationWarning: Gdk.Cursor.new is deprecated
cursors.MOVE : Gdk.Cursor.new(Gdk.CursorType.FLEUR),
/usr/local/lib/python3.5/dist-packages/matplotlib/backends/backend_gtk3agg.py:18: UserWarning: The Gtk3Agg backend is known to not work on Python 3.x with pycairo. Try installing cairocffi.
"The Gtk3Agg backend is known to not work on Python 3.x with pycairo. "
/usr/local/lib/python3.5/dist-packages/scipy/lib/decorator.py:219: DeprecationWarning: inspect.getargspec() is deprecated, use inspect.signature() instead
first = inspect.getargspec(caller)[0][0] # first arg
/usr/local/lib/python3.5/dist-packages/scipy/optimize/nonlin.py:1498: DeprecationWarning: inspect.getargspec() is deprecated, use inspect.signature() instead
args, varargs, varkw, defaults = inspect.getargspec(jac.init)
/usr/local/lib/python3.5/dist-packages/scipy/stats/_distn_infrastructure.py:611: DeprecationWarning: inspect.getargspec() is deprecated, use inspect.signature() instead
sign = inspect.getargspec(self._stats)
/usr/local/lib/python3.5/dist-packages/scipy/stats/_distn_infrastructure.py:648: DeprecationWarning: inspect.getargspec() is deprecated, use inspect.signature() instead
shapes_args = inspect.getargspec(meth)
****** WARNING: using empty weights ******
USB: Transferring Data...
[Error 25] Myriad Error: "5x5 Convolution outputChannels Dimension too small.".

Sign In or Register to comment.