dcgan.py

import torch.nn as nn


class DCGAN3D_G(nn.Module):
    def __init__(self, isize, nz, nc, ngf, ngpu, n_extra_layers=0):
        super(DCGAN3D_G, self).__init__()
        self.ngpu = ngpu
        assert isize % 16 == 0, "isize has to be a multiple of 16"

        cngf, tisize = ngf // 2, 4
        while tisize != isize:
            cngf = cngf * 2
            tisize = tisize * 2

        main = nn.Sequential(
            # input is Z, going into a convolution
            nn.ConvTranspose3d(nz, cngf, 4, 1, 0, bias=False),
            nn.BatchNorm3d(cngf),
            nn.ReLU(True),
        )

        i, csize, cndf = 3, 4, cngf
        while csize < isize // 2:
            main.add_module(str(i),
                            nn.ConvTranspose3d(cngf, cngf // 2, 4, 2, 1, bias=False))
            main.add_module(str(i + 1),
                            nn.BatchNorm3d(cngf // 2))
            main.add_module(str(i + 2),
                            nn.ReLU(True))
            i += 3
            cngf = cngf // 2
            csize = csize * 2

        # Extra layers
        for t in range(n_extra_layers):
            main.add_module(str(i),
                            nn.Conv3d(cngf, cngf, 3, 1, 1, bias=False))
            main.add_module(str(i + 1),
                            nn.BatchNorm3d(cngf))
            main.add_module(str(i + 2),
                            nn.ReLU(True))
            i += 3

        main.add_module(str(i),
                        nn.ConvTranspose3d(cngf, nc, 4, 2, 1, bias=False))
        main.add_module(str(i + 1), nn.Tanh())
        self.main = main

    def forward(self, input):
        return self.main(input)