Table 5.

Prediction performance of the suggested SER system for the IEMOCAP dataset using speech spectrograms.

Emotions/Classes	Recall Values	Precision Values	F1-Score
Anger	0.85	0.68	0.76
Sad	0.74	0.72	0.75
Happy	0.69	0.85	0.73
Neutral	0.81	0.78	0.80
Weighted Acc	0.77	0.76	0.77
Un-weighted Acc	0.76	0.76	0.76
Accuracy	-	-	0.77