niftynet.layer.downsample_res_block module

class DownBlock(n_output_chns=4, kernel_size=3, downsample_kernel_size=2, downsample_stride=2, acti_func='relu', w_initializer=None, w_regularizer=None, type_string='bn_acti_conv', name='res-downsample')

Bases: niftynet.layer.base_layer.TrainableLayer

layer_op(inputs, is_training=True)

Consists of:

(inputs)--conv_0-o-conv_1--conv_2-+-(conv_res)--down_sample--
                 |                |
                 o----------------o

conv_0, conv_res is also returned for feature forwarding purpose