niftynet.layer.squeeze_excitation module

class ChannelSELayer(func='AVG', reduction_ratio=16, name='channel_squeeze_excitation')[source]

Bases: niftynet.layer.base_layer.Layer

Re-implementation of Squeeze-and-Excitation (SE) block described in:

Hu et al., Squeeze-and-Excitation Networks, arXiv:1709.01507
layer_op(input_tensor)[source]
class SpatialSELayer(name='spatial_squeeze_excitation')[source]

Bases: niftynet.layer.base_layer.Layer

Re-implementation of SE block – squeezing spatially and exciting channel-wise described in:

Roy et al., Concurrent Spatial and Channel Squeeze & Excitation
in Fully Convolutional Networks, arXiv:1803.02579
layer_op(input_tensor)[source]
class ChannelSpatialSELayer(func='AVG', reduction_ratio=16, name='channel_spatial_squeeze_excitation')[source]

Bases: niftynet.layer.base_layer.Layer

Re-implementation of concurrent spatial and channel squeeze & excitation:

Roy et al., Concurrent Spatial and Channel Squeeze & Excitation
in Fully Convolutional Networks, arXiv:1803.02579
layer_op(input_tensor)[source]