Beyond Words: An Intelligent Human-Machine Dialogue System with Multimodal Generation and Emotional Comprehension

Table 3

In-depth analysis of different emotion recognition methods.

Method	Precision	Recall	F1
SVM-(Speech)	45.93	45.16	45.54
SVM-(Visual)	52.01	52.10	52.25
SVM-(Speech + Visual)	59.28	59.79	59.53
CNN + RNN-(Speech)	65.93	64.83	65.43
CNN + RNN-(Visual)	69.23	67.29	68.25
CNN + RNN-(Speech + Visual)	71.81	72.01	71.91

Bold values indicate the best results.