vgg16.py
Go to the documentation of this file.
00001 import chainer
00002 import chainer.functions as F
00003 import chainer.links as L
00004 
00005 
00006 class VGG16(chainer.Chain):
00007 
00008     def __init__(self, n_class):
00009         super(VGG16, self).__init__(
00010             conv1_1=L.Convolution2D(3, 64, 3, stride=1, pad=1),
00011             conv1_2=L.Convolution2D(64, 64, 3, stride=1, pad=1),
00012 
00013             conv2_1=L.Convolution2D(64, 128, 3, stride=1, pad=1),
00014             conv2_2=L.Convolution2D(128, 128, 3, stride=1, pad=1),
00015 
00016             conv3_1=L.Convolution2D(128, 256, 3, stride=1, pad=1),
00017             conv3_2=L.Convolution2D(256, 256, 3, stride=1, pad=1),
00018             conv3_3=L.Convolution2D(256, 256, 3, stride=1, pad=1),
00019 
00020             conv4_1=L.Convolution2D(256, 512, 3, stride=1, pad=1),
00021             conv4_2=L.Convolution2D(512, 512, 3, stride=1, pad=1),
00022             conv4_3=L.Convolution2D(512, 512, 3, stride=1, pad=1),
00023 
00024             conv5_1=L.Convolution2D(512, 512, 3, stride=1, pad=1),
00025             conv5_2=L.Convolution2D(512, 512, 3, stride=1, pad=1),
00026             conv5_3=L.Convolution2D(512, 512, 3, stride=1, pad=1),
00027 
00028             fc6=L.Linear(25088, 4096),
00029             fc7=L.Linear(4096, 4096),
00030             fc8=L.Linear(4096, n_class)
00031         )
00032 
00033     def __call__(self, x, t=None):
00034         h = F.relu(self.conv1_1(x))
00035         h = F.relu(self.conv1_2(h))
00036         h = F.max_pooling_2d(h, 2, stride=2)
00037 
00038         h = F.relu(self.conv2_1(h))
00039         h = F.relu(self.conv2_2(h))
00040         h = F.max_pooling_2d(h, 2, stride=2)
00041 
00042         h = F.relu(self.conv3_1(h))
00043         h = F.relu(self.conv3_2(h))
00044         h = F.relu(self.conv3_3(h))
00045         h = F.max_pooling_2d(h, 2, stride=2)
00046 
00047         h = F.relu(self.conv4_1(h))
00048         h = F.relu(self.conv4_2(h))
00049         h = F.relu(self.conv4_3(h))
00050         h = F.max_pooling_2d(h, 2, stride=2)
00051 
00052         h = F.relu(self.conv5_1(h))
00053         h = F.relu(self.conv5_2(h))
00054         h = F.relu(self.conv5_3(h))
00055         h = F.max_pooling_2d(h, 2, stride=2)
00056 
00057         h = F.dropout(F.relu(self.fc6(h)), ratio=0.5)
00058         h = F.dropout(F.relu(self.fc7(h)), ratio=0.5)
00059         h = self.fc8(h)
00060 
00061         self.pred = F.softmax(h)
00062         if t is None:
00063             assert not chainer.config.train
00064             return
00065 
00066         self.loss = F.softmax_cross_entropy(h, t)
00067         self.acc = F.accuracy(h, t)
00068         return self.loss


jsk_recognition_utils
Author(s):
autogenerated on Tue Jul 2 2019 19:40:37