def forward(self, x):
x = quant(x)
x = roundmax(x)
tmp = Variable(torch.zeros(1, 3, 32, 32).cuda())
f = fft.Fft2d()
fft_rout, fft_iout = f(x, tmp)
mag = torch.sqrt(
torch.mul(fft_rout, fft_rout) + torch.mul(fft_iout, fft_iout))
tmp = torch.zeros(1, 1, 32, 32).cuda()
tmp = torch.add(torch.add(mag[:, 0, :, :], mag[:, 1, :, :]),
mag[:, 2, :, :])
f = open(args.outputfile, 'a+')
for i in range(0, 32):
for j in range(0, 32):
print(tmp[0, i, j].data[0] / 3, file=f)
f.close()
'''
f = open('clean_train_freq.txt','w')
for i in range(0,32):
for j in range(0,32):
print(tmp.data[0,i,j]/3,file = f)
f.close()
'''
#print(mag.size())
#print(tmp.size())
#####################################
# print original image
'''
f = open('Fixedimage.txt','a+')
for i in range(0,32):
for j in range(0,32):
tmp2 = torch.add(torch.add(x[0,0,i,j],x[0,1,i,j]),x[0,2,i,j])
print(tmp2.data[0]/3,file = f, end='\t')
#print(mag[0,1,i,j].data[0],file = f)
#print(mag[0,2,i,j].data[0],file = f)
print('',file=f)
f.close()
'''
#####################################
#####################################
# print fft image
'''
f = open('Fixedfft.txt','a+')
for i in range(0,mag.size()[2]):
for j in range(0,mag.size()[3]):
print(tmp[0,i,j].data[0]/3,file = f, end='\t')
#print(mag[0,1,i,j].data[0],file = f)
#print(mag[0,2,i,j].data[0],file = f)
print('',file=f)
f.close()
'''
#####################################
'''
def forward(self, x):
x = quant(x)
x = roundmax(x)
tmp = Variable(torch.zeros(1, 3, 32, 32).cuda())
f = fft.Fft2d()
fft_rout, fft_iout = f(x, tmp)
mag = torch.sqrt(
torch.mul(fft_rout, fft_rout) + torch.mul(fft_iout, fft_iout))
tmp = torch.zeros(1, 1, 32, 32).cuda()
tmp = torch.add(torch.add(mag[:, 0, :, :], mag[:, 1, :, :]),
mag[:, 2, :, :])
tmp = torch.abs(tmp)
PFSUM = 0
'''
for i in range(0,32):
for j in range(0,32):
if (i+j) > 15:
print_value = 0
elif (i-j) < 17:
print_value = 0
elif (j-i) < 17:
print_value = 0
elif (i+j) < 48:
print_value = 0
else:
PFSUM = PFSUM + tmp[0,i,j]
'''
freq_l = 0
freq_h = 0
for i in range(0, 32):
for j in range(0, 32):
#'''
if (i + j) < 11:
freq_l = freq_l + torch.abs(tmp[0, i, j])
elif (i - j) > 21:
freq_l = freq_l + torch.abs(tmp[0, i, j])
elif (j - i) > 21:
freq_l = freq_l + torch.abs(tmp[0, i, j])
elif (i + j) > 52:
freq_l = freq_l + torch.abs(tmp[0, i, j])
else:
freq_h = freq_h + torch.abs(tmp[0, i, j])
PFSUM = freq_l - freq_h
f = open(args.outputfile, 'a+')
print(PFSUM.item(), file=f)
f.close()
'''
f = open('clean_train_freq.txt','w')
for i in range(0,32):
for j in range(0,32):
print(tmp.data[0,i,j]/3,file = f)
f.close()
'''
#print(mag.size())
#print(tmp.size())
#####################################
# print original image
'''
f = open('Fixedimage.txt','a+')
for i in range(0,32):
for j in range(0,32):
tmp2 = torch.add(torch.add(x[0,0,i,j],x[0,1,i,j]),x[0,2,i,j])
print(tmp2.data[0]/3,file = f, end='\t')
#print(mag[0,1,i,j].data[0],file = f)
#print(mag[0,2,i,j].data[0],file = f)
print('',file=f)
f.close()
'''
#####################################
#####################################
# print fft image
'''
f = open('Fixedfft.txt','a+')
for i in range(0,mag.size()[2]):
for j in range(0,mag.size()[3]):
print(tmp[0,i,j].data[0]/3,file = f, end='\t')
#print(mag[0,1,i,j].data[0],file = f)
#print(mag[0,2,i,j].data[0],file = f)
print('',file=f)
f.close()
'''
#####################################
'''
def forward(self, x):
if (args.gau == 0) & (args.blur == 0):
#no noise
pass
elif (args.blur == 0) & (args.gau != 0):
#gaussian noise add
gau_kernel = torch.randn(x.size()) * args.gau
x = Variable(gau_kernel.cuda()) + x
elif (args.gau == 0) & (args.blur != 0):
#blur noise add
blur_kernel_partial = torch.FloatTensor(
utils.genblurkernel(args.blur))
blur_kernel_partial = torch.matmul(
blur_kernel_partial.unsqueeze(1),
torch.transpose(blur_kernel_partial.unsqueeze(1), 0, 1))
kernel_size = blur_kernel_partial.size()[0]
zeros = torch.zeros(kernel_size, kernel_size)
blur_kernel = torch.cat(
(blur_kernel_partial, zeros, zeros, zeros, blur_kernel_partial,
zeros, zeros, zeros, blur_kernel_partial), 0)
blur_kernel = blur_kernel.view(3, 3, kernel_size, kernel_size)
blur_padding = int((blur_kernel_partial.size()[0] - 1) / 2)
#x = torch.nn.functional.conv2d(x, weight=blur_kernel.cuda(), padding=blur_padding)
x = torch.nn.functional.conv2d(x,
weight=Variable(blur_kernel.cuda()),
padding=blur_padding)
elif (args.gau != 0) & (args.blur != 0):
#both gaussian and blur noise added
blur_kernel_partial = torch.FloatTensor(
utils.genblurkernel(args.blur))
blur_kernel_partial = torch.matmul(
blur_kernel_partial.unsqueeze(1),
torch.transpose(blur_kernel_partial.unsqueeze(1), 0, 1))
kernel_size = blur_kernel_partial.size()[0]
zeros = torch.zeros(kernel_size, kernel_size)
blur_kernel = torch.cat(
(blur_kernel_partial, zeros, zeros, zeros, blur_kernel_partial,
zeros, zeros, zeros, blur_kernel_partial), 0)
blur_kernel = blur_kernel.view(3, 3, kernel_size, kernel_size)
blur_padding = int((blur_kernel_partial.size()[0] - 1) / 2)
x = torch.nn.functional.conv2d(x,
weight=Variable(blur_kernel.cuda()),
padding=blur_padding)
gau_kernel = torch.randn(x.size()) * args.gau
x = Variable(gau_kernel.cuda()) + x
else:
print("Something is wrong in noise adding part")
exit()
tmp = Variable(torch.zeros(1, 3, 224, 224).cuda())
f = fft.Fft2d()
fft_rout, fft_iout = f(x, tmp)
mag = torch.sqrt(
torch.mul(fft_rout, fft_rout) + torch.mul(fft_iout, fft_iout))
tmp = torch.zeros(1, 1, 224, 224).cuda()
tmp = torch.add(torch.add(mag[:, 0, :, :], mag[:, 1, :, :]),
mag[:, 2, :, :])
tmp = torch.abs(tmp)
PFSUM = 0
for i in range(0, 224):
for j in range(0, 224):
if (i + j) < 167:
print_value = 0
elif (i - j) > 56:
print_value = 0
elif (j - i) > 56:
print_value = 0
elif (i + j) > 279:
print_value = 0
else:
PFSUM = PFSUM + tmp[0, i, j]
f = open(args.outputfile, 'a+')
print(PFSUM.item(), file=f)
f.close()
#f = open(args.outputfile,'a+')
#try:
# print(PFSUM.item(),file=f)
#except:
# print(PFSUM,file=f)
#f.close()
'''
f = open(args.outputfile,'a+')
for i in range(0,224):
for j in range(0,224):
print(tmp[0,i,j].item()/3,file = f)
f.close()
exit()
'''
"""
if args.fixed:
x = quant(x)
x = roundmax(x)
out = self.conv1(x)
if args.fixed:
out = quant(out)
out = roundmax(out)
out = self.conv2(out)
if args.fixed:
out = quant(out)
out = roundmax(out)
out = self.conv3(out)
if args.fixed:
out = quant(out)
out = roundmax(out)
out = self.conv4(out)
if args.fixed:
out = quant(out)
out = roundmax(out)
out = self.conv5(out)
if args.fixed:
out = quant(out)
out = roundmax(out)
out = self.conv6(out)
if args.fixed:
out = quant(out)
out = roundmax(out)
out = self.conv7(out)
if args.fixed:
out = quant(out)
out = roundmax(out)
out = self.conv8(out)
if args.fixed:
out = quant(out)
out = roundmax(out)
out = self.conv9(out)
if args.fixed:
out = quant(out)
out = roundmax(out)
out = self.conv10(out)
if args.fixed:
out = quant(out)
out = roundmax(out)
out = self.conv11(out)
if args.fixed:
out = quant(out)
out = roundmax(out)
out = self.conv12(out)
if args.fixed:
out = quant(out)
out = roundmax(out)
out = self.conv13(out)
out = out.view(out.size(0), -1)
if args.fixed:
out = quant(out)
out = roundmax(out)
out = self.linear1(out)
if args.fixed:
out = quant(out)
out = roundmax(out)
out = self.linear2(out)
if args.fixed:
out = quant(out)
out = roundmax(out)
out = self.linear3(out)
if args.fixed:
out = quant(out)
out = roundmax(out)
"""
out = torch.zeros(1000)
return out