EURASIP Journal on Audio, Speech, and Music Processing

Table 1 Total number of parameters for DOANet models

From: DOANet: a deep dilated convolutional neural network approach for search and rescue with drone-embedded sound source localization

Model	Trainable	Non-trainable	Total
Azimuth	67,938,850	16,320	67,955,170
Elevation	34,902,177	11,200	34,913,377

Back to article page