Skip to main content
EURASIP Journal on Audio, Speech, and Music Processing
Fig. 5 | EURASIP Journal on Audio, Speech, and Music Processing

Fig. 5

From: Trainable windows for SincNet architecture

Fig. 5

Mean and standard-deviation of CER for all the investigated window functions a at minimum loss (left axis) and at 100th epoch (right axis). b Enlarged view of the CER at 100th epoch for Group1-A, Group1-B, and Group3 functions, and c Group2 functions

Back to article page