A Voice Cloning Method Based on the Improved HiFi-GAN Model

Figure 2

Network structure based on the x-vector extracted by the TDNN.