minor correction

perceivelab · Mar 15, 2021 · 4a42d23 · 4a42d23
1 parent bbb1a91
commit 4a42d23
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 26 deletions.
diff --git a/test_DSL.py b/test_DSL.py
@@ -12,13 +12,11 @@
             {'source': 'Hollywood', 'path': os.path.join('data','Hollywood2','test')},
             {'source': 'UCFSports', 'path': os.path.join('data','UCF','test')}]
 
-
 dataset_index = 0
 fromVideo=False
 dev = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
 image_size=(128, 192)
 
-
 dataset_source = source_datasets[dataset_index]['source']
 encoder_pretrained = False
 
@@ -76,7 +74,6 @@ def main():
     file_info.writelines(info)
     file_info.close()
 
-
     if fromVideo:
         if dataset_source=='LEDOV' or dataset_source=='UAV123':
             list_video= pd.read_csv(os.path.join('data',dataset_source,'test.csv'))['0'].values.tolist()
@@ -105,7 +102,7 @@ def main():
 
             original_length= len(list_frames)
 
-            #if number of video frames are less of 2*lentemporal, we append the frames to the list by going back
+            #if the number of video frames are less of 2*lentemporal, we append the frames to the list by going back
             if original_length<2*len_temporal-1:
                 num_missed_frames =  2*len_temporal -1 - original_length
                 for k in range(num_missed_frames):
@@ -122,7 +119,6 @@ def main():
                 for i in tqdm(range(len(list_frames))):
                     img = list_frames[i]
 
-
                     snippet.append(img)
 
                     if i<original_length:
@@ -172,9 +168,7 @@ def resized_frames_from_video(v, path_video):
     print(os.path.join(path_video,v))
     vidcap = cv2.VideoCapture(os.path.join(path_video,v))
     success,image = vidcap.read()
-
     frames=[]
-
     success = True
     while success:
         image = cv2.resize(image, dsize=(image_size[1], image_size[0]),interpolation=cv2.INTER_CUBIC)

diff --git a/train_DSL.py b/train_DSL.py
@@ -19,23 +19,14 @@
 from dataset.videoDataset import Dataset3D
 from dataset.infiniteDataLoader import InfiniteDataLoader
 
-'''
+
 source_datasets = [{'source': 'DHF1K', 'path': os.path.join('data','DHF1K','source')}, 
             {'source': 'Hollywood', 'path': os.path.join('data','Hollywood2','train')},
             {'source': 'UCFSports', 'path': os.path.join('data','UCF','train')}]
 
 validation_datasets = [{'source': 'DHF1K', 'path': os.path.join('data','DHF1K','source')},
             {'source': 'Hollywood', 'path': os.path.join('data','Hollywood2','train')},
             {'source': 'UCFSports', 'path': os.path.join('data','UCF','train')}]
-'''
-
-source_datasets = [{'source': 'DHF1K', 'path': os.path.join("C:\\","Users","gbellitto","Desktop","GitRepository","video-saliency-detection","data","DHF1K","source")}, 
-            {'source': 'Hollywood', 'path': os.path.join("C:\\","Users","gbellitto","Desktop","GitRepository","video-saliency-detection","data",'Hollywood2','train')},
-            {'source': 'UCFSports', 'path': os.path.join("C:\\","Users","gbellitto","Desktop","GitRepository","video-saliency-detection","data",'UCF','train')}]
-
-validation_datasets = [{'source': 'DHF1K', 'path': os.path.join("C:\\","Users","gbellitto","Desktop","GitRepository","video-saliency-detection","data",'DHF1K','source')},
-            {'source': 'Hollywood', 'path': os.path.join("C:\\","Users","gbellitto","Desktop","GitRepository","video-saliency-detection","data",'Hollywood2','train')},
-            {'source': 'UCFSports', 'path': os.path.join("C:\\","Users","gbellitto","Desktop","GitRepository","video-saliency-detection","data",'UCF','train')}]
 
 
 def main():
@@ -87,7 +78,6 @@ def main():
 
     for idx, p in enumerate(path_source_data):
 
-        #print(idx)
         print(p)
 
         if 'LEDOV' in p:
@@ -428,7 +418,7 @@ def main():
                 plt.savefig(os.path.join('output', subfolder, test_name,'loss_validation.png'))
                 plt.close()
 
-                #Plot validation loss Per-Dataset
+                #Plot validation per-dataset loss 
                 x = torch.arange(1, len(check_point['per_dataset_loss'][f'val_{list_source_validation[0]}'])+1).numpy()
                 for s in list_source_validation:
                     plt.plot(x, check_point['per_dataset_loss'][f'val_{s}'], label=f"val_loss {s}")
@@ -452,15 +442,10 @@ def main():
 
 
 def transform(snippet):
-    snippet = np.concatenate(snippet, axis=-1)  # 224 x 384 x 24
-    snippet = torch.from_numpy(snippet).permute(2, 0, 1).contiguous().float()  #24 x 224 x 384
+    snippet = np.concatenate(snippet, axis=-1)
+    snippet = torch.from_numpy(snippet).permute(2, 0, 1).contiguous().float()
     snippet = snippet.mul_(2.).sub_(255).div(255)
-                                # 1 x 8 x 3 x 224 x 384           1 x 3 x 8 x 224 x 384
-                                #          |                           |            
     snippet = snippet.view(1,-1,3,snippet.size(1),snippet.size(2)).permute(0,2,1,3,4) 
-    ''' a differenza del transform usato nel dataloader qui aggiungiuamo quell' 1 iniziale, a rappresentare 
-        la dimensione del batch, perchè il modello vuole in input: batch x C x len_temporal x H x W
-    '''
     return snippet   
 
 if __name__ == '__main__':