Name

dnn_pool - Pooling DNN activation for each channel.

Synopsis

dnn_pool -act Activation [-layer Layer [Layer …]] [-chn Channel [Channel …]] [-dmask DnnMask] -meth Method -out Output

Arguments

Required Arguments

Argument	Discription
act	Path of a .act.h5 file which contains activation information to extract features.
meth	Method of activation pooling.Choices are ‘max’, ‘mean’, ‘median’, ‘L1’, ‘L2’.The L1 means 1-norm and L2 means 2-norm.
out	Output path with a suffix as .act.h5.

Optional Arguments

Argument	Discription
layer	Name of the target layer(s).Default is all.E.g., ‘conv1’ represents the first convolution layer, and ‘fc1’ represents the first full connection layer.
chn	Index of target channel(s).Default is all.Channel index starts from 1.
dmask	Path of a .dmask.csv file in which detailed information of neuron(s) of interest in DNN is specified.Argument layer/chn and dmask are mutually exclusive. Provide only one of them if needed.

Outputs

An .act.h5 file containing the pooling activation that can be read and saved with the module dnnbrain.io.fileio.ActivationFile.

Examples

These examples demonstrate the activation pooling extraction function. Activation provided by test.act.h5 was pooling and finally saved in the dnn_fe.act.h5 file.

# Asserting target layers using the -layer argument
dnn_pool -act ./test.act.h5 -layer conv5 -meth max -out ./dnn_pool1.act.h5

# Asserting target layers using the -dmask argument
dnn_pool -act ./test.act.h5 -dmask ./test.dmask.csv -meth max -out ./dnn_pool2.act.h5

# Not asserting layers which means target layers and chns are all
dnn_pool -act ./test.act.h5 -meth max -out ./dnn_pool3.act.h5