EURASIP Journal on Advances in Signal Processing

Table 7 Comparison of different fusion strategies

From: The construction of Bert fusion model of speech recognition and sensing for South China electricity charge service scenario

Model selection	Accuracy	Recall	F1 value
Text CNN	0.8833	0.8826	0.8826
Bi LSTM	0.8646	0.8635	0.8644
SA-BLTCNN	0.9184	0.9164	0.9124
Transformer	0.8791	0.8787	0.8783
Simple voting-MIX	0.9215	0.9231	0.9225
Stacking-MIX	0.9283	0.9275	0.9277
Adaboost-MIX	0.9284	0.9299	0.9292

Back to article page