downsampled_logits_shape[0],
img_shape[1],
img_shape[2],
downsampled_logits_shape[3]
])
pool4_feature = end_points['vgg_16/pool4']
with tf.variable_scope('vgg_16/fc8'):
aux_logits_16s = slim.conv2d(pool4_feature, number_of_classes, [1, 1],
activation_fn=None,
weights_initializer=tf.zeros_initializer,
scope='conv_pool4')
# Perform the upsampling
upsample_filter_np_x2 = bilinear_upsample_weights(2, # upsample_factor,
number_of_classes)
upsample_filter_tensor_x2 = tf.Variable(upsample_filter_np_x2, name='vgg_16/fc8/t_conv_x2')
upsampled_logits = tf.nn.conv2d_transpose(logits, upsample_filter_tensor_x2,
output_shape=tf.shape(aux_logits_16s),
strides=[1, 2, 2, 1],
padding='SAME')
upsampled_logits = upsampled_logits + aux_logits_16s
upsample_filter_np_x16 = bilinear_upsample_weights(upsample_factor,
number_of_classes)
upsample_filter_tensor_x16 = tf.Variable(upsample_filter_np_x16, name='vgg_16/fc8/t_conv_x16')
downsampled_logits_shape[0],
img_shape[1],
img_shape[2],
downsampled_logits_shape[3]
])
pool4_feature = end_points['vgg_16/pool4']
with tf.variable_scope('vgg_16/fc8'):
aux_logits_16s = slim.conv2d(pool4_feature, number_of_classes, [1, 1],
activation_fn=None,
weights_initializer=tf.zeros_initializer,
scope='conv_pool4')
# Perform the upsampling
upsample_filter_np_x2 = bilinear_upsample_weights(2, # upsample_factor,
number_of_classes)
upsample_filter_tensor_x2 = tf.Variable(upsample_filter_np_x2, name='vgg_16/fc8/t_conv_x2')
upsampled_logits = tf.nn.conv2d_transpose(logits, upsample_filter_tensor_x2,
output_shape=tf.shape(aux_logits_16s),
strides=[1, 2, 2, 1],
padding='SAME')
upsampled_logits = upsampled_logits + aux_logits_16s
############################ Added Code Here ########################
# Add the x2 upsampleing
pool3_feature = end_points['vgg_16/pool3']
with tf.variable_scope('vgg_16/fc8'):
aux_logits_8s = slim.conv2d(pool3_feature, number_of_classes, [1, 1],
activation_fn=None,
weights_initializer=tf.zeros_initializer,
scope='conv_pool3')
# pool4 prediction
pool4_feature = end_points['vgg_16/pool4']
with tf.variable_scope('vgg_16/fc8'):
aux_logits_pool4 = slim.conv2d(pool4_feature,
number_of_classes, [1, 1],
activation_fn=None,
weights_initializer=tf.zeros_initializer,
scope='conv_pool4')
# Perform the upsampling
upsamples_stride_2 = 2
upsample_filter_np_2x = bilinear_upsample_weights(upsamples_stride_2,
number_of_classes)
upsample_filter_tensor_2x = tf.Variable(upsample_filter_np_2x,
name='vgg_16/fc8/t_conv_2x')
upsampled_logits = tf.nn.conv2d_transpose(
logits,
upsample_filter_tensor_2x,
output_shape=tf.shape(aux_logits_pool4),
strides=[1, upsamples_stride_2, upsamples_stride_2, 1],
padding='SAME')
# Combining
upsampled_logits = upsampled_logits + aux_logits_pool4
# Combining predictions from both the final layer and the pool4 layer, at stride 2
# upsamples_stride_2
upsample_filter_np_2x_2 = bilinear_upsample_weights(upsamples_stride_2,
number_of_classes)
