Generative Adversarial Networks Projects

The architecture of the discriminator network

The discriminator network contains five volumetric convolutional layers with the following configuration:

• 3D convolutional layers: 5
• Channels: 64, 128, 256, 512, 1
• Kernel sizes: 4, 4, 4, 4, 4
• Strides: 2, 2, 2, 2, 1
• Activations: Leaky ReLU, Leaky ReLU, Leaky ReLU, Leaky ReLU, Sigmoid
• Batch normalization: Yes, Yes, Yes, Yes, None
• Pooling layers: No, No, No, No, No
• Linear layers: No, No, No, No, No

The input and output of the network are as follows:

• Input: A 3D image with shape (64, 64, 64) 
• Output: The probability of the input data belonging to either the real or the fake class

The flow of the tensors and the input and output shapes of the tensors for each layer in the discriminator network ...