port to python3

.gitignore

@@ -0,0 +1,3 @@
+*.pth
+*.pyc
+launch.json

model/PieAPPv0pt1_PT.py

@@ -33,7 +33,7 @@ def __init__(self,batch_size,num_patches):
 		self.num_patches = num_patches	
 
 	def flatten(self,matrix): # takes NxCxHxW input and outputs NxHWC
-		return matrix.view((self.batch_size*self.num_patches,-1))
+		return matrix.reshape(int(self.batch_size*self.num_patches),-1)
 
 	def compute_features(self,input):
 		#conv1 -> relu -> conv2 -> relu -> pool2 -> conv3 -> relu
@@ -57,12 +57,12 @@ def compute_score(self,image_A_patches, image_ref_patches):
 		diff_coarse = ref_coarse - A_coarse		
 		# per patch score: fc1_score -> relu -> fc2_score
 		per_patch_score = self.ref_score_subtract(0.01*self.fc2_score(F.relu(self.fc1_score(diff_ms))))
-		per_patch_score.view((-1,self.num_patches))
+		per_patch_score.reshape(-1,int(self.num_patches))
 		# per patch weight: fc1_weight -> relu -> fc2_weight
 		const = Variable(torch.from_numpy(0.000001*np.ones((1,))).float(), requires_grad=False) 
 		const_cuda = const.cuda()		
 		per_patch_weight = self.fc2_weight(F.relu(self.fc1_weight(diff_coarse)))+const_cuda
-		per_patch_weight.view((-1,self.num_patches))
+		per_patch_weight.reshape(-1,int(self.num_patches))
 		product_val = torch.mul(per_patch_weight,per_patch_score)
 		dot_product_val = torch.sum(product_val)
 		norm_factor = torch.sum(per_patch_weight)

test_PieAPP_PT.py

@@ -12,10 +12,10 @@
 
 ######## check for model and download if not present
 if not os.path.isfile('weights/PieAPPv0.1.pth'):
-	print "downloading dataset"
+	print("downloading dataset")
 	os.system("bash scripts/download_PieAPPv0.1_PT_weights.sh")
 	if not os.path.isfile('weights/PieAPPv0.1.pth'):
-		print "PieAPPv0.1.pth not downloaded"
+		print("PieAPPv0.1.pth not downloaded")
 		sys.exit()
 
 ######## variables
@@ -58,7 +58,10 @@
 
 ######## initialize the model
 PieAPP_net = PieAPP(batch_size,num_patches_per_dim)
-PieAPP_net.load_state_dict(torch.load('weights/PieAPPv0.1.pth'))
+weights = torch.load('weights/PieAPPv0.1.pth')
+# modify ref_score_subtract.weight so instead of have torch.Size([1]) it has torch.Size([1, 1])
+weights['ref_score_subtract.weight'] = weights['ref_score_subtract.weight'].unsqueeze(0)
+PieAPP_net.load_state_dict(weights)
 
 if use_gpu == 1:
 	PieAPP_net.cuda()
@@ -67,8 +70,8 @@
 weight_accum = 0.0
 
 # iterate through smaller size sub-images (to prevent memory overload)
-for x_iter in range(0, -(-num_x//num_patches)):	
-	for y_iter in range(0, -(-num_y//num_patches)):
+for x_iter in range(0, -(-num_x//num_patches_per_dim)):	
+	for y_iter in range(0, -(-num_y//num_patches_per_dim)):
 		# compute the size of the subimage
 		if (num_patches_per_dim*(x_iter + 1) >= num_x):				
 			size_slice_cols = cols - x_loc[num_patches_per_dim*x_iter]
@@ -100,4 +103,4 @@
 		score_accum += np.sum(np.multiply(curr_err, curr_weights))
 		weight_accum += np.sum(curr_weights)
 
-print 'PieAPP value of '+args.A_path+ ' with respect to: '+str(score_accum/weight_accum)
+print('PieAPP value of '+args.A_path+ ' with respect to: '+str(score_accum/weight_accum))