Emotional Dialogue Generation Based on Conditional Variational Autoencoder and Dual Emotion Framework

Figure 3

Transformer Encoder for feature extraction.