pyramid-flow / video_vae /modeling_discriminator.py
multimodalart's picture
Upload 33 files
f0533a5 verified
import functools
import torch.nn as nn
from einops import rearrange
import torch
def weights_init(m):
classname = m.__class__.__name__
if classname.find('Conv') != -1:
nn.init.normal_(m.weight.data, 0.0, 0.02)
nn.init.constant_(m.bias.data, 0)
elif classname.find('BatchNorm') != -1:
nn.init.normal_(m.weight.data, 1.0, 0.02)
nn.init.constant_(m.bias.data, 0)
class NLayerDiscriminator(nn.Module):
"""Defines a PatchGAN discriminator as in Pix2Pix
--> see https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix/blob/master/models/networks.py
"""
def __init__(self, input_nc=3, ndf=64, n_layers=4):
"""Construct a PatchGAN discriminator
Parameters:
input_nc (int) -- the number of channels in input images
ndf (int) -- the number of filters in the last conv layer
n_layers (int) -- the number of conv layers in the discriminator
norm_layer -- normalization layer
"""
super(NLayerDiscriminator, self).__init__()
# norm_layer = nn.BatchNorm2d
norm_layer = nn.InstanceNorm2d
if type(norm_layer) == functools.partial: # no need to use bias as BatchNorm2d has affine parameters
use_bias = norm_layer.func != nn.BatchNorm2d
else:
use_bias = norm_layer != nn.BatchNorm2d
kw = 4
padw = 1
sequence = [nn.Conv2d(input_nc, ndf, kernel_size=kw, stride=2, padding=padw), nn.LeakyReLU(0.2, True)]
nf_mult = 1
nf_mult_prev = 1
for n in range(1, n_layers): # gradually increase the number of filters
nf_mult_prev = nf_mult
nf_mult = min(2 ** n, 8)
sequence += [
nn.Conv2d(ndf * nf_mult_prev, ndf * nf_mult, kernel_size=kw, stride=2, padding=padw, bias=use_bias),
norm_layer(ndf * nf_mult),
nn.LeakyReLU(0.2, True)
]
nf_mult_prev = nf_mult
nf_mult = min(2 ** n_layers, 8)
sequence += [
nn.Conv2d(ndf * nf_mult_prev, ndf * nf_mult, kernel_size=kw, stride=1, padding=padw, bias=use_bias),
norm_layer(ndf * nf_mult),
nn.LeakyReLU(0.2, True)
]
sequence += [
nn.Conv2d(ndf * nf_mult, 1, kernel_size=kw, stride=1, padding=padw)] # output 1 channel prediction map
self.main = nn.Sequential(*sequence)
def forward(self, input):
"""Standard forward."""
return self.main(input)
class NLayerDiscriminator3D(nn.Module):
"""Defines a 3D PatchGAN discriminator as in Pix2Pix but for 3D inputs."""
def __init__(self, input_nc=3, ndf=64, n_layers=3, use_actnorm=False):
"""
Construct a 3D PatchGAN discriminator
Parameters:
input_nc (int) -- the number of channels in input volumes
ndf (int) -- the number of filters in the last conv layer
n_layers (int) -- the number of conv layers in the discriminator
use_actnorm (bool) -- flag to use actnorm instead of batchnorm
"""
super(NLayerDiscriminator3D, self).__init__()
# if not use_actnorm:
# norm_layer = nn.BatchNorm3d
# else:
# raise NotImplementedError("Not implemented.")
norm_layer = nn.InstanceNorm3d
if type(norm_layer) == functools.partial:
use_bias = norm_layer.func != nn.BatchNorm3d
else:
use_bias = norm_layer != nn.BatchNorm3d
kw = 4
padw = 1
sequence = [nn.Conv3d(input_nc, ndf, kernel_size=kw, stride=2, padding=padw), nn.LeakyReLU(0.2, True)]
nf_mult = 1
nf_mult_prev = 1
for n in range(1, n_layers): # gradually increase the number of filters
nf_mult_prev = nf_mult
nf_mult = min(2 ** n, 8)
sequence += [
nn.Conv3d(ndf * nf_mult_prev, ndf * nf_mult, kernel_size=(kw, kw, kw), stride=(1,2,2), padding=padw, bias=use_bias),
norm_layer(ndf * nf_mult),
nn.LeakyReLU(0.2, True)
]
nf_mult_prev = nf_mult
nf_mult = min(2 ** n_layers, 8)
sequence += [
nn.Conv3d(ndf * nf_mult_prev, ndf * nf_mult, kernel_size=(kw, kw, kw), stride=1, padding=padw, bias=use_bias),
norm_layer(ndf * nf_mult),
nn.LeakyReLU(0.2, True)
]
sequence += [nn.Conv3d(ndf * nf_mult, 1, kernel_size=kw, stride=1, padding=padw)] # output 1 channel prediction map
self.main = nn.Sequential(*sequence)
def forward(self, input):
"""Standard forward."""
return self.main(input)