torchbenchmark/models/Background_Matting/networks.py

import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.nn.init as init
import numpy as np


class ResnetConditionHR(nn.Module):
	def __init__(self, input_nc, output_nc, ngf=64, nf_part=64,norm_layer=nn.BatchNorm2d, use_dropout=False, n_blocks1=7, n_blocks2=3, padding_type='reflect'):
		assert(n_blocks1 >= 0); assert(n_blocks2 >= 0)
		super(ResnetConditionHR, self).__init__()
		self.input_nc = input_nc
		self.output_nc = output_nc
		self.ngf = ngf
		use_bias=True

		#main encoder output 256xW/4xH/4
		model_enc1 = [nn.ReflectionPad2d(3),nn.Conv2d(input_nc[0], ngf, kernel_size=7, padding=0,bias=use_bias),norm_layer(ngf),nn.ReLU(True)]
		model_enc1 += [nn.Conv2d(ngf , ngf * 2, kernel_size=3,stride=2, padding=1, bias=use_bias),norm_layer(ngf * 2),nn.ReLU(True)]
		model_enc2 = [nn.Conv2d(ngf*2 , ngf * 4, kernel_size=3,stride=2, padding=1, bias=use_bias),norm_layer(ngf * 4),nn.ReLU(True)]
		

		#back encoder output 256xW/4xH/4
		model_enc_back = [nn.ReflectionPad2d(3),nn.Conv2d(input_nc[1], ngf, kernel_size=7, padding=0,bias=use_bias),norm_layer(ngf),nn.ReLU(True)]
		n_downsampling = 2
		for i in range(n_downsampling):
			mult = 2**i
			model_enc_back += [nn.Conv2d(ngf * mult, ngf * mult * 2, kernel_size=3,stride=2, padding=1, bias=use_bias),norm_layer(ngf * mult * 2),nn.ReLU(True)]

		#seg encoder output 256xW/4xH/4
		model_enc_seg = [nn.ReflectionPad2d(3),nn.Conv2d(input_nc[2], ngf, kernel_size=7, padding=0,bias=use_bias),norm_layer(ngf),nn.ReLU(True)]
		n_downsampling = 2
		for i in range(n_downsampling):
			mult = 2**i
			model_enc_seg += [nn.Conv2d(ngf * mult, ngf * mult * 2, kernel_size=3,stride=2, padding=1, bias=use_bias),norm_layer(ngf * mult * 2),nn.ReLU(True)]

		mult = 2**n_downsampling

		# #motion encoder output 256xW/4xH/4
		model_enc_multi = [nn.ReflectionPad2d(3),nn.Conv2d(input_nc[3], ngf, kernel_size=7, padding=0,bias=use_bias),norm_layer(ngf),nn.ReLU(True)]
		n_downsampling = 2
		for i in range(n_downsampling):
			mult = 2**i
			model_enc_multi += [nn.Conv2d(ngf * mult, ngf * mult * 2, kernel_size=3,stride=2, padding=1, bias=use_bias),norm_layer(ngf * mult * 2),nn.ReLU(True)]


		self.model_enc1 = nn.Sequential(*model_enc1)
		self.model_enc2 = nn.Sequential(*model_enc2)
		self.model_enc_back = nn.Sequential(*model_enc_back)
		self.model_enc_seg = nn.Sequential(*model_enc_seg)
		self.model_enc_multi = nn.Sequential(*model_enc_multi)

		mult = 2**n_downsampling
		self.comb_back=nn.Sequential(nn.Conv2d(ngf * mult*2,nf_part,kernel_size=1,stride=1,padding=0,bias=False),norm_layer(ngf),nn.ReLU(True))
		self.comb_seg=nn.Sequential(nn.Conv2d(ngf * mult*2,nf_part,kernel_size=1,stride=1,padding=0,bias=False),norm_layer(ngf),nn.ReLU(True))
		self.comb_multi=nn.Sequential(nn.Conv2d(ngf * mult*2,nf_part,kernel_size=1,stride=1,padding=0,bias=False),norm_layer(ngf),nn.ReLU(True))

		#decoder
		model_res_dec=[nn.Conv2d(ngf * mult +3*nf_part,ngf*mult,kernel_size=1,stride=1,padding=0,bias=False),norm_layer(ngf*mult),nn.ReLU(True)]
		for i in range(n_blocks1):
			model_res_dec += [ResnetBlock(ngf * mult, padding_type=padding_type, norm_layer=norm_layer, use_dropout=use_dropout, use_bias=use_bias)]

		model_res_dec_al=[]
		for i in range(n_blocks2):
			model_res_dec_al += [ResnetBlock(ngf * mult, padding_type=padding_type, norm_layer=norm_layer, use_dropout=use_dropout, use_bias=use_bias)]

		model_res_dec_fg=[]
		for i in range(n_blocks2):
			model_res_dec_fg += [ResnetBlock(ngf * mult, padding_type=padding_type, norm_layer=norm_layer, use_dropout=use_dropout, use_bias=use_bias)]

		model_dec_al=[]
		for i in range(n_downsampling):
			mult = 2**(n_downsampling - i)
			#model_dec_al += [nn.ConvTranspose2d(ngf * mult, int(ngf * mult / 2),kernel_size=3, stride=2,padding=1, output_padding=1,bias=use_bias),norm_layer(int(ngf * mult / 2)),nn.ReLU(True)]
			model_dec_al += [nn.Upsample(scale_factor=2,mode='bilinear',align_corners = True),nn.Conv2d(ngf * mult, int(ngf * mult / 2), 3, stride=1,padding=1),norm_layer(int(ngf * mult / 2)),nn.ReLU(True)]
		model_dec_al += [nn.ReflectionPad2d(3),nn.Conv2d(ngf, 1, kernel_size=7, padding=0),nn.Tanh()]


		model_dec_fg1=[nn.Upsample(scale_factor=2,mode='bilinear',align_corners = True),nn.Conv2d(ngf * 4, int(ngf * 2), 3, stride=1,padding=1),norm_layer(int(ngf * 2)),nn.ReLU(True)]
		model_dec_fg2=[nn.Upsample(scale_factor=2,mode='bilinear',align_corners = True),nn.Conv2d(ngf * 4, ngf, 3, stride=1,padding=1),norm_layer(ngf),nn.ReLU(True),nn.ReflectionPad2d(3),nn.Conv2d(ngf, output_nc-1, kernel_size=7, padding=0)]

		self.model_res_dec = nn.Sequential(*model_res_dec)
		self.model_res_dec_al=nn.Sequential(*model_res_dec_al)
		self.model_res_dec_fg=nn.Sequential(*model_res_dec_fg)
		self.model_al_out=nn.Sequential(*model_dec_al)

		self.model_dec_fg1=nn.Sequential(*model_dec_fg1)
		self.model_fg_out = nn.Sequential(*model_dec_fg2)
		

	def forward(self, image,back,seg,multi):
		img_feat1=self.model_enc1(image)
		img_feat=self.model_enc2(img_feat1)

		back_feat=self.model_enc_back(back)
		seg_feat=self.model_enc_seg(seg)
		multi_feat=self.model_enc_multi(multi)

		oth_feat=torch.cat([self.comb_back(torch.cat([img_feat,back_feat],dim=1)),self.comb_seg(torch.cat([img_feat,seg_feat],dim=1)),self.comb_multi(torch.cat([img_feat,back_feat],dim=1))],dim=1)

		out_dec=self.model_res_dec(torch.cat([img_feat,oth_feat],dim=1))

		out_dec_al=self.model_res_dec_al(out_dec)
		al_out=self.model_al_out(out_dec_al)

		out_dec_fg=self.model_res_dec_fg(out_dec)
		out_dec_fg1=self.model_dec_fg1(out_dec_fg)
		fg_out=self.model_fg_out(torch.cat([out_dec_fg1,img_feat1],dim=1))


		return al_out, fg_out

############################## part ##################################



def conv_init(m):
	classname = m.__class__.__name__
	if classname.find('Conv') != -1:
		init.xavier_uniform(m.weight, gain=np.sqrt(2))
		#init.normal(m.weight)
		if m.bias is not None:
			init.constant(m.bias, 0)

	if classname.find('Linear') != -1:
		init.normal(m.weight)
		init.constant(m.bias,1)

	if classname.find('BatchNorm2d') != -1:
		init.normal(m.weight.data, 1.0, 0.2)
		init.constant(m.bias.data, 0.0)

class conv3x3(nn.Module):
	'''(conv => BN => ReLU)'''
	def __init__(self, in_ch, out_ch):
		super(conv3x3, self).__init__()
		self.conv = nn.Sequential(
			nn.Conv2d(in_ch, out_ch, 3, stride=2,padding=1),
			nn.BatchNorm2d(out_ch),
			nn.LeakyReLU(0.2,inplace=True),
		)

	def forward(self, x):
		x = self.conv(x)
		return x

class conv3x3s1(nn.Module):
	'''(conv => BN => ReLU)'''
	def __init__(self, in_ch, out_ch):
		super(conv3x3s1, self).__init__()
		self.conv = nn.Sequential(
			nn.Conv2d(in_ch, out_ch, 3, stride=1,padding=1),
			nn.BatchNorm2d(out_ch),
			nn.LeakyReLU(0.2,inplace=True),
		)

	def forward(self, x):
		x = self.conv(x)
		return x




class conv1x1(nn.Module):
	'''(conv => BN => ReLU)'''
	def __init__(self, in_ch, out_ch):
		super(conv1x1, self).__init__()
		self.conv = nn.Sequential(
			nn.Conv2d(in_ch, out_ch, 1, stride=1,padding=0),
			nn.BatchNorm2d(out_ch),
			nn.LeakyReLU(0.2,inplace=True),
		)

	def forward(self, x):
		x = self.conv(x)
		return x



class upconv3x3(nn.Module):
	def __init__(self, in_ch, out_ch):
		super(upconv3x3, self).__init__()
		self.conv = nn.Sequential(
			nn.Upsample(scale_factor=2,mode='bilinear'),
			nn.Conv2d(in_ch, out_ch, 3, stride=1,padding=1),
			nn.BatchNorm2d(out_ch),
			nn.ReLU(inplace=True),
		)
	def forward(self, x):
		x=self.conv(x)
		return x

class fc(nn.Module):
	def __init__(self,in_ch,out_ch):
		super(fc,self).__init__()
		self.fullc = nn.Sequential(
			nn.Linear(in_ch,out_ch),
			nn.ReLU(inplace=True),
		)
	def forward(self,x):
		x=self.fullc(x)
		return x

# Define a resnet block
class ResnetBlock(nn.Module):
	def __init__(self, dim, padding_type, norm_layer, use_dropout, use_bias):
		super(ResnetBlock, self).__init__()
		self.conv_block = self.build_conv_block(dim, padding_type, norm_layer, use_dropout, use_bias)

	def build_conv_block(self, dim, padding_type, norm_layer, use_dropout, use_bias):
		conv_block = []
		p = 0
		if padding_type == 'reflect':
			conv_block += [nn.ReflectionPad2d(1)]
		elif padding_type == 'replicate':
			conv_block += [nn.ReplicationPad2d(1)]
		elif padding_type == 'zero':
			p = 1
		else:
			raise NotImplementedError('padding [%s] is not implemented' % padding_type)

		conv_block += [nn.Conv2d(dim, dim, kernel_size=3, padding=p, bias=use_bias),
					   norm_layer(dim),
					   nn.ReLU(True)]
		if use_dropout:
			conv_block += [nn.Dropout(0.5)]

		p = 0
		if padding_type == 'reflect':
			conv_block += [nn.ReflectionPad2d(1)]
		elif padding_type == 'replicate':
			conv_block += [nn.ReplicationPad2d(1)]
		elif padding_type == 'zero':
			p = 1
		else:
			raise NotImplementedError('padding [%s] is not implemented' % padding_type)
		conv_block += [nn.Conv2d(dim, dim, kernel_size=3, padding=p, bias=use_bias),
					   norm_layer(dim)]

		return nn.Sequential(*conv_block)

	def forward(self, x):
		out = x + self.conv_block(x)
		return out


##################################### Discriminators ####################################################

class MultiscaleDiscriminator(nn.Module):
	def __init__(self, input_nc, ndf=64, n_layers=3, norm_layer=nn.BatchNorm2d, 
				 use_sigmoid=False, num_D=3, getIntermFeat=False):
		super(MultiscaleDiscriminator, self).__init__()
		self.num_D = num_D
		self.n_layers = n_layers
		self.getIntermFeat = getIntermFeat
	 
		for i in range(num_D):
			netD = NLayerDiscriminator(input_nc, ndf, n_layers, norm_layer, use_sigmoid, getIntermFeat)
			if getIntermFeat:                                
				for j in range(n_layers+2):
					setattr(self, 'scale'+str(i)+'_layer'+str(j), getattr(netD, 'model'+str(j)))                                   
			else:
				setattr(self, 'layer'+str(i), netD.model)

		self.downsample = nn.AvgPool2d(3, stride=2, padding=[1, 1], count_include_pad=False)

	def singleD_forward(self, model, input):
		if self.getIntermFeat:
			result = [input]
			for i in range(len(model)):
				result.append(model[i](result[-1]))
			return result[1:]
		else:
			return [model(input)]

	def forward(self, input):        
		num_D = self.num_D
		result = []
		input_downsampled = input
		for i in range(num_D):
			if self.getIntermFeat:
				model = [getattr(self, 'scale'+str(num_D-1-i)+'_layer'+str(j)) for j in range(self.n_layers+2)]
			else:
				model = getattr(self, 'layer'+str(num_D-1-i))
			result.append(self.singleD_forward(model, input_downsampled))
			if i != (num_D-1):
				input_downsampled = self.downsample(input_downsampled)
		return result
		
# Defines the PatchGAN discriminator with the specified arguments.
class NLayerDiscriminator(nn.Module):
	def __init__(self, input_nc, ndf=64, n_layers=3, norm_layer=nn.BatchNorm2d, use_sigmoid=False, getIntermFeat=False):
		super(NLayerDiscriminator, self).__init__()
		self.getIntermFeat = getIntermFeat
		self.n_layers = n_layers

		kw = 4
		padw = int(np.ceil((kw-1.0)/2))
		sequence = [[nn.Conv2d(input_nc, ndf, kernel_size=kw, stride=2, padding=padw), nn.LeakyReLU(0.2, True)]]

		nf = ndf
		for n in range(1, n_layers):
			nf_prev = nf
			nf = min(nf * 2, 512)
			sequence += [[
				nn.Conv2d(nf_prev, nf, kernel_size=kw, stride=2, padding=padw),
				norm_layer(nf), nn.LeakyReLU(0.2, True)
			]]

		nf_prev = nf
		nf = min(nf * 2, 512)
		sequence += [[
			nn.Conv2d(nf_prev, nf, kernel_size=kw, stride=1, padding=padw),
			norm_layer(nf),
			nn.LeakyReLU(0.2, True)
		]]

		sequence += [[nn.Conv2d(nf, 1, kernel_size=kw, stride=1, padding=padw)]]

		if use_sigmoid:
			sequence += [[nn.Sigmoid()]]

		if getIntermFeat:
			for n in range(len(sequence)):
				setattr(self, 'model'+str(n), nn.Sequential(*sequence[n]))
		else:
			sequence_stream = []
			for n in range(len(sequence)):
				sequence_stream += sequence[n]
			self.model = nn.Sequential(*sequence_stream)

	def forward(self, input):
		if self.getIntermFeat:
			res = [input]
			for n in range(self.n_layers+2):
				model = getattr(self, 'model'+str(n))
				res.append(model(res[-1]))
			return res[1:]
		else:
			return self.model(input)