Two-Level Multimodal Fusion for Sentiment Analysis in Public Security

Table 1

Comparison experimental results for sentiment analysis on the CMU-MOSI and CMU-MOSEI datasets.

Metric	Acc-2	F-score	MAE	Acc-7	Corr
Case of the CMU-MOSI dataset
TFN	80.82	80.77	0.901	23.94	0.698
LMF	82.53	82.47	0.917	33.23	0.695
MuLT	79.11	79.10	0.977	33.23	0.666
TlMF	92.28	92.26	0.373	65.49	0.933
Case of the CMU-MOSI dataset
TFN	78.57	78.09	0.623	47.21	0.648
LMF	78.03	78.18	0.609	48.02	0.657
MuLT	79.35	79.29	0.631	48.45	0.647
TlMF	79.46	79.46	0.645	48.88	0.669