Merge pull request #18 from adynathos/master

Adjustments for Python 3

datasets/__init__.py

@@ -1,2 +1,2 @@
-import cityscapes
-import voc
+from . import cityscapes
+from . import voc

datasets/cityscapes.py

@@ -53,7 +53,7 @@ class CityScapes(data.Dataset):
     def __init__(self, quality, mode, joint_transform=None, transform=None, target_transform=None):
         self.imgs = make_dataset(quality, mode)
         if len(self.imgs) == 0:
-            raise (RuntimeError('Found 0 images, please check the data set'))
+            raise RuntimeError('Found 0 images, please check the data set')
         self.quality = quality
         self.mode = mode
         self.joint_transform = joint_transform
@@ -72,7 +72,7 @@ def __getitem__(self, index):
 
         mask = np.array(mask)
         mask_copy = mask.copy()
-        for k, v in self.id_to_trainid.iteritems():
+        for k, v in self.id_to_trainid.items():
             mask_copy[mask == k] = v
         mask = Image.fromarray(mask_copy.astype(np.uint8))
 

datasets/voc.py

@@ -64,7 +64,7 @@ class VOC(data.Dataset):
     def __init__(self, mode, joint_transform=None, transform=None, target_transform=None):
         self.imgs = make_dataset(mode)
         if len(self.imgs) == 0:
-            raise (RuntimeError('Found 0 images, please check the data set'))
+            raise RuntimeError('Found 0 images, please check the data set')
         self.mode = mode
         self.joint_transform = joint_transform
         self.transform = transform

eval/eval_voc.py

@@ -21,7 +21,7 @@
 
 def main():
     net = PSPNet(num_classes=voc.num_classes).cuda()
-    print 'load model ' + args['snapshot']
+    print('load model ' + args['snapshot'])
     net.load_state_dict(torch.load(os.path.join(ckpt_path, args['exp_name'], args['snapshot'])))
     net.eval()
 
@@ -48,7 +48,7 @@ def main():
         prediction = voc.colorize_mask(prediction)
         prediction.save(os.path.join(ckpt_path, args['exp_name'], 'test', img_name + '.png'))
 
-        print '%d / %d' % (vi + 1, len(test_loader))
+        print('%d / %d' % (vi + 1, len(test_loader)))
 
 
 if __name__ == '__main__':

models/fcn16s.py

@@ -2,7 +2,7 @@
 from torch import nn
 from torchvision import models
 
-from utils import get_upsampling_weight
+from ..utils import get_upsampling_weight
 from .config import vgg16_caffe_path
 
 

models/fcn32s.py

@@ -2,7 +2,7 @@
 from torch import nn
 from torchvision import models
 
-from utils import get_upsampling_weight
+from ..utils import get_upsampling_weight
 from .config import vgg16_caffe_path
 
 

models/fcn8s.py

@@ -2,7 +2,7 @@
 from torch import nn
 from torchvision import models
 
-from utils import get_upsampling_weight
+from ..utils import get_upsampling_weight
 from .config import vgg16_path, vgg16_caffe_path
 
 

models/gcn.py

@@ -3,7 +3,7 @@
 from torch import nn
 from torchvision import models
 
-from utils import initialize_weights
+from ..utils import initialize_weights
 from .config import res152_path
 
 

models/psp_net.py

@@ -3,8 +3,8 @@
 from torch import nn
 from torchvision import models
 
-from utils import initialize_weights
-from utils.misc import Conv2dDeformable
+from ..utils import initialize_weights
+from ..utils.misc import Conv2dDeformable
 from .config import res101_path
 
 

models/seg_net.py

@@ -2,7 +2,7 @@
 from torch import nn
 from torchvision import models
 
-from utils import initialize_weights
+from ..utils import initialize_weights
 from .config import vgg19_bn_path
 
 

models/u_net.py

@@ -2,7 +2,7 @@
 import torch.nn.functional as F
 from torch import nn
 
-from utils import initialize_weights
+from ..utils import initialize_weights
 
 
 class _EncoderBlock(nn.Module):

train/cityscapes-fcn (caffe vgg)/train.py

@@ -47,7 +47,7 @@ def main():
         curr_epoch = 1
         args['best_record'] = {'epoch': 0, 'val_loss': 1e10, 'acc': 0, 'acc_cls': 0, 'mean_iu': 0, 'fwavacc': 0}
     else:
-        print 'training resumes from ' + args['snapshot']
+        print('training resumes from ' + args['snapshot'])
         net.load_state_dict(torch.load(os.path.join(ckpt_path, exp_name, args['snapshot'])))
         split_snapshot = args['snapshot'].split('_')
         curr_epoch = int(split_snapshot[1]) + 1
@@ -141,8 +141,8 @@ def train(train_loader, net, criterion, optimizer, epoch, train_args):
         writer.add_scalar('train_loss', train_loss.avg, curr_iter)
 
         if (i + 1) % train_args['print_freq'] == 0:
-            print '[epoch %d], [iter %d / %d], [train loss %.5f]' % (
-                epoch, i + 1, len(train_loader), train_loss.avg)
+            print('[epoch %d], [iter %d / %d], [train loss %.5f]' % (
+                epoch, i + 1, len(train_loader), train_loss.avg))
 
 
 def validate(val_loader, net, criterion, optimizer, epoch, train_args, restore, visualize):
@@ -209,15 +209,15 @@ def validate(val_loader, net, criterion, optimizer, epoch, train_args, restore,
         val_visual = vutils.make_grid(val_visual, nrow=3, padding=5)
         writer.add_image(snapshot_name, val_visual)
 
-        print '-----------------------------------------------------------------------------------------------------------'
-    print '[epoch %d], [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f]' % (
-        epoch, val_loss.avg, acc, acc_cls, mean_iu, fwavacc)
+        print('-----------------------------------------------------------------------------------------------------------')
+    print('[epoch %d], [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f]' % (
+        epoch, val_loss.avg, acc, acc_cls, mean_iu, fwavacc))
 
-    print 'best record: [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f], [epoch %d]' % (
+    print('best record: [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f], [epoch %d]' % (
         train_args['best_record']['val_loss'], train_args['best_record']['acc'], train_args['best_record']['acc_cls'],
-        train_args['best_record']['mean_iu'], train_args['best_record']['fwavacc'], train_args['best_record']['epoch'])
+        train_args['best_record']['mean_iu'], train_args['best_record']['fwavacc'], train_args['best_record']['epoch']))
 
-    print '-----------------------------------------------------------------------------------------------------------'
+    print('-----------------------------------------------------------------------------------------------------------')
 
     writer.add_scalar('val_loss', val_loss.avg, epoch)
     writer.add_scalar('acc', acc, epoch)

train/cityscapes-fcn/train.py

@@ -47,7 +47,7 @@ def main():
         curr_epoch = 1
         args['best_record'] = {'epoch': 0, 'val_loss': 1e10, 'acc': 0, 'acc_cls': 0, 'mean_iu': 0, 'fwavacc': 0}
     else:
-        print 'training resumes from ' + args['snapshot']
+        print('training resumes from ' + args['snapshot'])
         net.load_state_dict(torch.load(os.path.join(ckpt_path, exp_name, args['snapshot'])))
         split_snapshot = args['snapshot'].split('_')
         curr_epoch = int(split_snapshot[1]) + 1
@@ -135,8 +135,8 @@ def train(train_loader, net, criterion, optimizer, epoch, train_args):
         writer.add_scalar('train_loss', train_loss.avg, curr_iter)
 
         if (i + 1) % train_args['print_freq'] == 0:
-            print '[epoch %d], [iter %d / %d], [train loss %.5f]' % (
-                epoch, i + 1, len(train_loader), train_loss.avg)
+            print('[epoch %d], [iter %d / %d], [train loss %.5f]' % (
+                epoch, i + 1, len(train_loader), train_loss.avg))
 
 
 def validate(val_loader, net, criterion, optimizer, epoch, train_args, restore, visualize):
@@ -203,15 +203,15 @@ def validate(val_loader, net, criterion, optimizer, epoch, train_args, restore,
         val_visual = vutils.make_grid(val_visual, nrow=3, padding=5)
         writer.add_image(snapshot_name, val_visual)
 
-    print '-----------------------------------------------------------------------------------------------------------'
-    print '[epoch %d], [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f]' % (
-        epoch, val_loss.avg, acc, acc_cls, mean_iu, fwavacc)
+    print('-----------------------------------------------------------------------------------------------------------')
+    print('[epoch %d], [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f]' % (
+        epoch, val_loss.avg, acc, acc_cls, mean_iu, fwavacc))
 
-    print 'best record: [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f], [epoch %d]' % (
+    print('best record: [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f], [epoch %d]' % (
         train_args['best_record']['val_loss'], train_args['best_record']['acc'], train_args['best_record']['acc_cls'],
-        train_args['best_record']['mean_iu'], train_args['best_record']['fwavacc'], train_args['best_record']['epoch'])
+        train_args['best_record']['mean_iu'], train_args['best_record']['fwavacc'], train_args['best_record']['epoch']))
 
-    print '-----------------------------------------------------------------------------------------------------------'
+    print('-----------------------------------------------------------------------------------------------------------')
 
     writer.add_scalar('val_loss', val_loss.avg, epoch)
     writer.add_scalar('acc', acc, epoch)

train/cityscapes-psp_net/train.py

@@ -48,7 +48,7 @@ def main():
         curr_epoch = 1
         args['best_record'] = {'epoch': 0, 'val_loss': 1e10, 'acc': 0, 'acc_cls': 0, 'mean_iu': 0, 'fwavacc': 0}
     else:
-        print 'training resumes from ' + args['snapshot']
+        print('training resumes from ' + args['snapshot'])
         net.load_state_dict(torch.load(os.path.join(ckpt_path, exp_name, args['snapshot'])))
         split_snapshot = args['snapshot'].split('_')
         curr_epoch = int(split_snapshot[1]) + 1
@@ -156,9 +156,9 @@ def train(train_loader, net, criterion, optimizer, curr_epoch, train_args, val_l
             writer.add_scalar('lr', optimizer.param_groups[1]['lr'], curr_iter)
 
             if (i + 1) % train_args['print_freq'] == 0:
-                print '[epoch %d], [iter %d / %d], [train main loss %.5f], [train aux loss %.5f]. [lr %.10f]' % (
+                print('[epoch %d], [iter %d / %d], [train main loss %.5f], [train aux loss %.5f]. [lr %.10f]' % (
                     curr_epoch, i + 1, len(train_loader), train_main_loss.avg, train_aux_loss.avg,
-                    optimizer.param_groups[1]['lr'])
+                    optimizer.param_groups[1]['lr']))
             if curr_iter >= train_args['max_iter']:
                 return
         validate(val_loader, net, criterion, optimizer, curr_epoch, train_args, restore, visualize)
@@ -190,7 +190,7 @@ def validate(val_loader, net, criterion, optimizer, epoch, train_args, restore,
                 inputs_all.append(None)
         gts_all[vi, :, :] = gts.data.squeeze_(0).cpu().numpy()
 
-        print 'validating: %d / %d' % (vi + 1, len(val_loader))
+        print('validating: %d / %d' % (vi + 1, len(val_loader)))
 
     acc, acc_cls, mean_iu, fwavacc = evaluate(predictions_all, gts_all, cityscapes.num_classes)
 
@@ -227,15 +227,15 @@ def validate(val_loader, net, criterion, optimizer, epoch, train_args, restore,
         val_visual = vutils.make_grid(val_visual, nrow=3, padding=5)
         writer.add_image(snapshot_name, val_visual)
 
-    print '-----------------------------------------------------------------------------------------------------------'
-    print '[epoch %d], [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f]' % (
-        epoch, val_loss.avg, acc, acc_cls, mean_iu, fwavacc)
+    print('-----------------------------------------------------------------------------------------------------------')
+    print('[epoch %d], [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f]' % (
+        epoch, val_loss.avg, acc, acc_cls, mean_iu, fwavacc))
 
-    print 'best record: [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f], [epoch %d]' % (
+    print('best record: [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f], [epoch %d]' % (
         train_args['best_record']['val_loss'], train_args['best_record']['acc'], train_args['best_record']['acc_cls'],
-        train_args['best_record']['mean_iu'], train_args['best_record']['fwavacc'], train_args['best_record']['epoch'])
+        train_args['best_record']['mean_iu'], train_args['best_record']['fwavacc'], train_args['best_record']['epoch']))
 
-    print '-----------------------------------------------------------------------------------------------------------'
+    print('-----------------------------------------------------------------------------------------------------------')
 
     writer.add_scalar('val_loss', val_loss.avg, epoch)
     writer.add_scalar('acc', acc, epoch)

train/voc-fcn (caffe vgg)/train.py

@@ -42,7 +42,7 @@ def main(train_args):
         curr_epoch = 1
         train_args['best_record'] = {'epoch': 0, 'val_loss': 1e10, 'acc': 0, 'acc_cls': 0, 'mean_iu': 0, 'fwavacc': 0}
     else:
-        print 'training resumes from ' + train_args['snapshot']
+        print('training resumes from ' + train_args['snapshot'])
         net.load_state_dict(torch.load(os.path.join(ckpt_path, exp_name, train_args['snapshot'])))
         split_snapshot = train_args['snapshot'].split('_')
         curr_epoch = int(split_snapshot[1]) + 1
@@ -128,9 +128,9 @@ def train(train_loader, net, criterion, optimizer, epoch, train_args):
         writer.add_scalar('train_loss', train_loss.avg, curr_iter)
 
         if (i + 1) % train_args['print_freq'] == 0:
-            print '[epoch %d], [iter %d / %d], [train loss %.5f]' % (
+            print('[epoch %d], [iter %d / %d], [train loss %.5f]' % (
                 epoch, i + 1, len(train_loader), train_loss.avg
-            )
+            ))
 
 
 def validate(val_loader, net, criterion, optimizer, epoch, train_args, restore, visualize):
@@ -193,15 +193,15 @@ def validate(val_loader, net, criterion, optimizer, epoch, train_args, restore,
         val_visual = vutils.make_grid(val_visual, nrow=3, padding=5)
         writer.add_image(snapshot_name, val_visual)
 
-    print '--------------------------------------------------------------------'
-    print '[epoch %d], [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f]' % (
-        epoch, val_loss.avg, acc, acc_cls, mean_iu, fwavacc)
+    print('--------------------------------------------------------------------')
+    print('[epoch %d], [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f]' % (
+        epoch, val_loss.avg, acc, acc_cls, mean_iu, fwavacc))
 
-    print 'best record: [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f], [epoch %d]' % (
+    print('best record: [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f], [epoch %d]' % (
         train_args['best_record']['val_loss'], train_args['best_record']['acc'], train_args['best_record']['acc_cls'],
-        train_args['best_record']['mean_iu'], train_args['best_record']['fwavacc'], train_args['best_record']['epoch'])
+        train_args['best_record']['mean_iu'], train_args['best_record']['fwavacc'], train_args['best_record']['epoch']))
 
-    print '--------------------------------------------------------------------'
+    print('--------------------------------------------------------------------')
 
     writer.add_scalar('val_loss', val_loss.avg, epoch)
     writer.add_scalar('acc', acc, epoch)

train/voc-fcn/train.py

@@ -42,7 +42,7 @@ def main(train_args):
         curr_epoch = 1
         train_args['best_record'] = {'epoch': 0, 'val_loss': 1e10, 'acc': 0, 'acc_cls': 0, 'mean_iu': 0, 'fwavacc': 0}
     else:
-        print 'training resumes from ' + train_args['snapshot']
+        print('training resumes from ' + train_args['snapshot'])
         net.load_state_dict(torch.load(os.path.join(ckpt_path, exp_name, train_args['snapshot'])))
         split_snapshot = train_args['snapshot'].split('_')
         curr_epoch = int(split_snapshot[1]) + 1
@@ -124,9 +124,9 @@ def train(train_loader, net, criterion, optimizer, epoch, train_args):
         writer.add_scalar('train_loss', train_loss.avg, curr_iter)
 
         if (i + 1) % train_args['print_freq'] == 0:
-            print '[epoch %d], [iter %d / %d], [train loss %.5f]' % (
+            print('[epoch %d], [iter %d / %d], [train loss %.5f]' % (
                 epoch, i + 1, len(train_loader), train_loss.avg
-            )
+            ))
 
 
 def validate(val_loader, net, criterion, optimizer, epoch, train_args, restore, visualize):
@@ -189,15 +189,15 @@ def validate(val_loader, net, criterion, optimizer, epoch, train_args, restore,
         val_visual = vutils.make_grid(val_visual, nrow=3, padding=5)
         writer.add_image(snapshot_name, val_visual)
 
-    print '--------------------------------------------------------------------'
-    print '[epoch %d], [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f]' % (
-        epoch, val_loss.avg, acc, acc_cls, mean_iu, fwavacc)
+    print('--------------------------------------------------------------------')
+    print('[epoch %d], [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f]' % (
+        epoch, val_loss.avg, acc, acc_cls, mean_iu, fwavacc))
 
-    print 'best record: [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f], [epoch %d]' % (
+    print('best record: [val loss %.5f], [acc %.5f], [acc_cls %.5f], [mean_iu %.5f], [fwavacc %.5f], [epoch %d]' % (
         train_args['best_record']['val_loss'], train_args['best_record']['acc'], train_args['best_record']['acc_cls'],
-        train_args['best_record']['mean_iu'], train_args['best_record']['fwavacc'], train_args['best_record']['epoch'])
+        train_args['best_record']['mean_iu'], train_args['best_record']['fwavacc'], train_args['best_record']['epoch']))
 
-    print '--------------------------------------------------------------------'
+    print('--------------------------------------------------------------------')
 
     writer.add_scalar('val_loss', val_loss.avg, epoch)
     writer.add_scalar('acc', acc, epoch)