Audio-Textual Emotion Recognition Based on Improved Neural Networks

Figure 4

The structure of the CBLA model for acoustics feature learning.