Research Article
Application of Generative Adversarial Nets (GANs) in Active Sound Production System of Electric Automobiles
Table 3
Generator structure of the spectrogram GAN model.
| | Layer | Type | Length and width of convolution kernels | Number of convolution kernels |
| | Input layer | Input layer | — | — | | Fully connected layer | Fully connected layer | 1∗100 | 16384 | | ReLU | Activation layer | — | — | | Trans Conv2D (stride = 2) | Convolutional layer | 5∗5 | 648 | | ReLU | Activation layer | — | — | | Trans Conv2D (stride = 2) | Convolutional layer | 5∗5 | 644 | | ReLU | Activation layer | — | — | | Trans Conv2D (stride = 2) | Convolutional layer | 5∗5 | 642 | | ReLU | Activation layer | — | — | | Trans Conv2D (stride = 2) | Convolutional layer | 5∗5 | 641 | | ReLU | Activation layer | — | — | | Trans Conv2D (stride = 2) | Convolutional layer | 5∗5 | 1 | | Tanh | Activation layer | — | — |
|
|
