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

<div>Overview of CVAE-DE. The blue part represents a Seq2Seq model based on the attention mechanism, which is the basis of our model. The orange and gold parts represent the conditional variational autoencoder model. The green part represents the dual emotion frame.</div>

Wireless Communications and Mobile Computing

fig1

Figure 1

Figure 1: Emotional Dialogue Generation Based on Conditional Variational Autoencoder and Dual Emotion Framework 