das.models_legacyΒΆ

Defines the network architectures.

Functions

cnn(nb_freq, nb_classes[, nb_channels, ...])

CNN for single-frequency and multi-channel data - uses 1D convolutions.

cnn2D(nb_freq, nb_classes[, nb_channels, ...])

CNN for multi-frequency and multi-channel data - uses 2D convolutions.

fcn(nb_freq, nb_classes[, nb_channels, ...])

[summary]

fcn2D(nb_freq, nb_classes[, nb_channels, ...])

[summary]

tcn(*args, **kwargs)

Synonym for tcn_stft.

tcn_multi(nb_freq, nb_classes[, nb_hist, ...])

Create TCN network with TCN layer as pre-processing and downsampling frontend with weights shared between channels.

tcn_seq(nb_freq, nb_classes[, nb_hist, ...])

Create TCN network.

tcn_small(nb_freq, nb_classes[, nb_hist, ...])

Create TCN network with TCN layer as pre-processing and downsampling frontend.

tcn_stft(nb_freq, nb_classes[, nb_hist, ...])

Create TCN network with optional trainable STFT layer as pre-processing and downsampling frontend.

tcn_tcn(nb_freq, nb_classes[, nb_hist, ...])

Create TCN network with TCN layer as pre-processing and downsampling frontend.