update list of files with errors in the padchest dataset (#137)

mlmed · Jun 20, 2023 · 23b1aad · 23b1aad
1 parent 8ea630c
commit 23b1aad
Showing 1 changed file with 12 additions and 1 deletion.
diff --git a/torchxrayvision/datasets.py b/torchxrayvision/datasets.py
@@ -894,7 +894,18 @@ def __init__(self,
                    "216840111366964012283393834152009033102258826_00-059-087.png",
                    "216840111366964012373310883942009170084120009_00-097-074.png",
                    "216840111366964012819207061112010315104455352_04-024-184.png",
-                   "216840111366964012819207061112010306085429121_04-020-102.png"]
+                   "216840111366964012819207061112010306085429121_04-020-102.png",
+                   "216840111366964012989926673512011083134050913_00-168-009.png", # broken PNG file (chunk b'\x00\x00\x00\x00')
+                   "216840111366964012373310883942009152114636712_00-102-045.png", # "OSError: image file is truncated"
+                   "216840111366964012819207061112010281134410801_00-129-131.png", # "OSError: image file is truncated"
+                   "216840111366964012487858717522009280135853083_00-075-001.png", # "OSError: image file is truncated"
+                   "216840111366964012989926673512011151082430686_00-157-045.png", # broken PNG file (chunk b'\x00\x00\x00\x00')
+                   "216840111366964013686042548532013208193054515_02-026-007.png", # "OSError: image file is truncated"
+                   "216840111366964013590140476722013058110301622_02-056-111.png", # "OSError: image file is truncated"
+                   "216840111366964013590140476722013043111952381_02-065-198.png", # "OSError: image file is truncated"
+                   "216840111366964013829543166512013353113303615_02-092-190.png", # "OSError: image file is truncated"
+                   "216840111366964013962490064942014134093945580_01-178-104.png", # "OSError: image file is truncated"
+                  ]
         self.csv = self.csv[~self.csv["ImageID"].isin(missing)]
 
         if unique_patients: