-
-
Save hma02/35cc44e5b17750453b1f to your computer and use it in GitHub Desktop.
Inception module in Lasagne (without 3x3s1 pooling)
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import lasagne as nn | |
Conv2DLayer = nn.layers.Conv2DDNNLayer | |
def inception_module(l_in, num_1x1, reduce_3x3, num_3x3, reduce_5x5, num_5x5, gain=1.0, bias=0.1): | |
""" | |
inception module (without the 3x3s1 pooling and projection because that's difficult in Theano right now) | |
""" | |
shape = l_in.get_output_shape() | |
out_layers = [] | |
# 1x1 | |
if num_1x1 > 0: | |
l_1x1 = nn.layers.NINLayer(l_in, num_units=num_1x1, W=nn.init.Orthogonal(gain), b=nn.init.Constant(bias)) | |
out_layers.append(l_1x1) | |
# 3x3 | |
if num_3x3 > 0: | |
if reduce_3x3 > 0: | |
l_reduce_3x3 = nn.layers.NINLayer(l_in, num_units=reduce_3x3, W=nn.init.Orthogonal(gain), b=nn.init.Constant(bias)) | |
else: | |
l_reduce_3x3 = l_in | |
l_3x3 = Conv2DLayer(l_reduce_3x3, num_filters=num_3x3, filter_size=(3, 3), border_mode="same", W=nn.init.Orthogonal(gain), b=nn.init.Constant(bias)) | |
out_layers.append(l_3x3) | |
# 5x5 | |
if num_5x5 > 0: | |
if reduce_5x5 > 0: | |
l_reduce_5x5 = nn.layers.NINLayer(l_in, num_units=reduce_5x5, W=nn.init.Orthogonal(gain), b=nn.init.Constant(bias)) | |
else: | |
l_reduce_5x5 = l_in | |
l_5x5 = Conv2DLayer(l_reduce_5x5, num_filters=num_5x5, filter_size=(5, 5), border_mode="same", W=nn.init.Orthogonal(gain), b=nn.init.Constant(bias)) | |
out_layers.append(l_5x5) | |
# stack | |
l_out = nn.layers.concat(out_layers) | |
return l_out |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment