def face_parse(image, label):
conv = dict(kernel=3, padding='SAME')
deconv = dict(kernel=3, padding='SAME', bias=True)
subpixel = dict(kernel=3, factor=2, padding='SAME')
with tf.default_args(conv=conv, deconv=deconv, subpixel=subpixel):
net = image
net = net.conv(16).bn().relu().conv(16).bn().relu().maxpool()
net = net.conv(32).bn().relu().conv(32).bn().relu().maxpool()
net = net.conv(64).bn().relu().conv(64).bn().relu().maxpool()
net = net.conv(64).bn().relu().conv(64).bn().relu().maxpool()
net = net.conv(64).bn().relu()
net = net.deconv(64).bn().relu()
net = net.subpixel().deconv(64).bn().relu().deconv(64).bn().relu()
net = net.subpixel().deconv(64).bn().relu().deconv(64).bn().relu()
net = net.subpixel().deconv(32).bn().relu().deconv(32).bn().relu()
net = net.subpixel().deconv(16).bn().relu().deconv(11, bias=True)
prob = net.softmax()
summary_parse(prob)
losses = tf.softmax_cross_entropy(net, label, name='losses')
loss = losses.mean()
return tf.dic(losses=losses, loss=loss, outputs=[prob])
def generator(z):
deconv = dict(kernel=4, stride=2, padding='SAME')
with tf.default_args(deconv=deconv):
net = z
net = net.dense(1024).bn().pleaky()
net = net.dense(7 * 7 * 128).bn().pleaky()
net = net.reshape((-1, 7, 7, 128))
net = net.deconv(64).bn().pleaky()
net = net.deconv(1).sigmoid()
return tf.summary_image(net, name='fake')
def discriminator(imgz):
conv = dict(kernel=4, stride=2, padding='SAME')
with tf.default_args(conv=conv):
net = imgz
net = net.conv(64).leaky()
net = net.conv(128).leaky().flat2d()
net = net.dense(1024).leaky()
net = net.dense(1).squeeze()
# prob = net.sigmoid()
# must return logits
return tf.identity(net, name='disc')
def discriminator(xx, num_cont, batch):
conv = dict(kernel=4, stride=2, bias=True, padding='SAME')
with tf.default_args(conv=conv):
net = xx
net = net.conv(64).leaky()
net = net.conv(128).leaky().flat2d()
net = net.dense(1024).leaky()
disc = net.dense(1).squeeze()
net = net.dense(128).leaky()
klass = net.dense(10)
cont = net[batch:].dense(num_cont).sigmoid()
return disc, klass, cont
def face_parse2(image, label):
conv = dict(kernel=3, padding='SAME')
deconv = dict(kernel=3, padding='SAME', bias=True)
# subpixel = dict(kernel=3, factor=2, padding='SAME')
maxpool_where = dict(kernel=2)
unpool_where = dict(kernel=2)
with tf.default_args(conv=conv,
deconv=deconv,
maxpool_where=maxpool_where,
unpool_where=unpool_where):
net = image
wheres = []
net, where = net.conv(16).bn().relu().conv(
16).bn().relu().maxpool_where()
wheres.append(where)
net, where = net.conv(32).bn().relu().conv(
32).bn().relu().maxpool_where()
wheres.append(where)
net, where = net.conv(64).bn().relu().conv(
64).bn().relu().maxpool_where()
wheres.append(where)
net, where = net.conv(64).bn().relu().conv(
64).bn().relu().maxpool_where()
wheres.append(where)
net = net.conv(64).bn().relu()
net = net.deconv(64).bn().relu()
net = net.unpool_where(
wheres.pop()).deconv(64).bn().relu().deconv(64).bn().relu()
net = net.unpool_where(
wheres.pop()).deconv(64).bn().relu().deconv(32).bn().relu()
net = net.unpool_where(
wheres.pop()).deconv(32).bn().relu().deconv(16).bn().relu()
net = net.unpool_where(wheres.pop()).deconv(16).bn().relu().deconv(
11, bias=True)
net = tf.summary_activation(net, name='logits')
prob = net.softmax()
summary_parse(prob)
# losses = tf.nn.sigmoid_cross_entropy_with_logits(net, label)
losses = tf.softmax_cross_entropy(net, label, name='losses')
loss = losses.mean()
return tf.dic(loss=loss, logits=net, label=label, image=image)