Adaptive representations of sound for automatic insect recognition
Table 5
Test and validation scores for the trained models using the leafFB frontend.
The median as well as the lower and upper limits are reported from training three runs of the same model with different randomization seeds and four convolutional layers.