{ "id": 33021, "title": "Corrupted image?", "url": "/competitions/intel-mobileodt-cervical-cancer-screening/discussion/33021", "author_name": "", "post_date": "2017-05-15T16:40:25.156929200Z", "votes": null, "comment_count": 4, "views": 0, "content": "

I try this image from additional:

\n\n
additional/Type_1/3068.jpg\n
\n\n

I try to load this with keras image preprocessor.

\n\n
[oystein@jupiter ]$ python\nPython 3.6.1 (default, Mar 27 2017, 00:27:06) \n[GCC 6.3.1 20170306] on linux\nType \"help\", \"copyright\", \"credits\" or \"license\" for more information.\n>>> from keras.preprocessing import image\nUsing Theano backend.\n>>> img = image.load_img(\"./additional/Type_1/3068.jpg\", target_size=(224, 224))\nTraceback (most recent call last):\n  File \"<stdin>\", line 1, in <module>\n  File \"/usr/lib/python3.6/site-packages/keras/preprocessing/image.py\", line 330, in load_img\n    img = img.resize(hw_tuple)\n  File \"/usr/lib/python3.6/site-packages/PIL/Image.py\", line 1630, in resize self.load()\n  File \"/usr/lib/python3.6/site-packages/PIL/ImageFile.py\", line 231, in load\n    \"(%d bytes not processed)\" % len(b))\nOSError: image file is truncated (107 bytes not processed)\n
\n\n

Has anyone else had the same problem with this image. Can anyone verify my md5sum of this image? 0b8cfe3f0d6256532dade1c1216024d8

\n\n

Update: There is another thread describing this. I should have checked the archive before posting.

", "messages": [ { "id": "182733", "postDate": "05/15/2017 16:40:25", "content": "

I try this image from additional:

\n\n
additional/Type_1/3068.jpg\n
\n\n

I try to load this with keras image preprocessor.

\n\n
[oystein@jupiter ]$ python\nPython 3.6.1 (default, Mar 27 2017, 00:27:06) \n[GCC 6.3.1 20170306] on linux\nType \"help\", \"copyright\", \"credits\" or \"license\" for more information.\n>>> from keras.preprocessing import image\nUsing Theano backend.\n>>> img = image.load_img(\"./additional/Type_1/3068.jpg\", target_size=(224, 224))\nTraceback (most recent call last):\n  File \"<stdin>\", line 1, in <module>\n  File \"/usr/lib/python3.6/site-packages/keras/preprocessing/image.py\", line 330, in load_img\n    img = img.resize(hw_tuple)\n  File \"/usr/lib/python3.6/site-packages/PIL/Image.py\", line 1630, in resize self.load()\n  File \"/usr/lib/python3.6/site-packages/PIL/ImageFile.py\", line 231, in load\n    \"(%d bytes not processed)\" % len(b))\nOSError: image file is truncated (107 bytes not processed)\n
\n\n

Has anyone else had the same problem with this image. Can anyone verify my md5sum of this image? 0b8cfe3f0d6256532dade1c1216024d8

\n\n

Update: There is another thread describing this. I should have checked the archive before posting.

", "rawMarkdown": "I try this image from additional:\n\n additional/Type_1/3068.jpg\n\nI try to load this with keras image preprocessor.\n\n [oystein@jupiter ]$ python\n Python 3.6.1 (default, Mar 27 2017, 00:27:06) \n [GCC 6.3.1 20170306] on linux\n Type \"help\", \"copyright\", \"credits\" or \"license\" for more information.\n >>> from keras.preprocessing import image\n Using Theano backend.\n >>> img = image.load_img(\"./additional/Type_1/3068.jpg\", target_size=(224, 224))\n Traceback (most recent call last):\n File \"", "votes": null }, { "id": "183225", "postDate": "05/17/2017 11:35:59", "content": "

Had same problem, but did know with which file. Now at least I know ;)

\n\n

ValueError: output of generator should be a tuple (x, y, sample_weight) or (x, y). Found: None

", "rawMarkdown": "Had same problem, but did know with which file. Now at least I know ;) \n\nValueError: output of generator should be a tuple (x, y, sample_weight) or (x, y). Found: None", "votes": null }, { "id": "183233", "postDate": "05/17/2017 12:46:57", "content": "

Thanks for your confirmation: Here are my corrupted(?) images:

\n\n
/additional/Type_1/3068.jpg\n/additional/Type_1/5893.jpg (empty)\n/additional/Type_2/2845.jpg (empty)\n/additional/Type_2/5892.jpg (empty)\n/additional/Type_2/7.jpg \n/train/Type_1/1339.jpg\n
\n\n

Of these six files, three of them is empty. That's not a mystery. The other three files are all JFIF standard 1.01, but there is also other files in this standard that loads perfectly. I guess I can open these three files in GIMP or something similar and save them in another standard. That may help.

", "rawMarkdown": "Thanks for your confirmation: Here are my corrupted(?) images:\n\n /additional/Type_1/3068.jpg\n /additional/Type_1/5893.jpg (empty)\n /additional/Type_2/2845.jpg (empty)\n /additional/Type_2/5892.jpg (empty)\n /additional/Type_2/7.jpg \n /train/Type_1/1339.jpg\n\nOf these six files, three of them is empty. That's not a mystery. The other three files are all JFIF standard 1.01, but there is also other files in this standard that loads perfectly. I guess I can open these three files in GIMP or something similar and save them in another standard. That may help.", "votes": null }, { "id": "183546", "postDate": "05/18/2017 14:19:51", "content": "

I haven't dug into the additional data, but I have the same problem with 1339.jpg. I was wondering if my download was corrupted, but if others are seeing the same thing there is probably a problem with the original data source.

", "rawMarkdown": "I haven't dug into the additional data, but I have the same problem with 1339.jpg. I was wondering if my download was corrupted, but if others are seeing the same thing there is probably a problem with the original data source.", "votes": null }, { "id": "183598", "postDate": "05/18/2017 17:30:58", "content": "

The MD5 hash of additional/Type_1/3068.jpg is 0B8CFE3F0D6256532DADE1C1216024D8 it's a valid file.\nThose are empty on my side too :\nadditional/Type_1/5893.jpg\nadditional/Type_2/2845.jpg\nadditional/Type_2/5892.jpg\nOthers are corrupted but I can still read them (partially) with C++ OpenCV.

", "rawMarkdown": "The MD5 hash of additional/Type_1/3068.jpg is 0B8CFE3F0D6256532DADE1C1216024D8 it's a valid file.\nThose are empty on my side too :\nadditional/Type_1/5893.jpg\nadditional/Type_2/2845.jpg\nadditional/Type_2/5892.jpg\nOthers are corrupted but I can still read them (partially) with C++ OpenCV.", "votes": null } ], "comments": [ { "id": 183225, "author_name": "matias3", "author_url": "", "post_date": "05/17/2017 11:35:59", "content": "

Had same problem, but did know with which file. Now at least I know ;)

\n\n

ValueError: output of generator should be a tuple (x, y, sample_weight) or (x, y). Found: None

", "votes": null, "replies": [ { "id": 183233, "author_name": "oysteijo", "author_url": "", "post_date": "05/17/2017 12:46:57", "content": "

Thanks for your confirmation: Here are my corrupted(?) images:

\n\n
/additional/Type_1/3068.jpg\n/additional/Type_1/5893.jpg (empty)\n/additional/Type_2/2845.jpg (empty)\n/additional/Type_2/5892.jpg (empty)\n/additional/Type_2/7.jpg \n/train/Type_1/1339.jpg\n
\n\n

Of these six files, three of them is empty. That's not a mystery. The other three files are all JFIF standard 1.01, but there is also other files in this standard that loads perfectly. I guess I can open these three files in GIMP or something similar and save them in another standard. That may help.

", "votes": null, "replies": [] }, { "id": 183546, "author_name": "gkericks", "author_url": "", "post_date": "05/18/2017 14:19:51", "content": "

I haven't dug into the additional data, but I have the same problem with 1339.jpg. I was wondering if my download was corrupted, but if others are seeing the same thing there is probably a problem with the original data source.

", "votes": null, "replies": [] } ] }, { "id": 183598, "author_name": "atmarouane", "author_url": "", "post_date": "05/18/2017 17:30:58", "content": "

The MD5 hash of additional/Type_1/3068.jpg is 0B8CFE3F0D6256532DADE1C1216024D8 it's a valid file.\nThose are empty on my side too :\nadditional/Type_1/5893.jpg\nadditional/Type_2/2845.jpg\nadditional/Type_2/5892.jpg\nOthers are corrupted but I can still read them (partially) with C++ OpenCV.

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "182733": "I try this image from additional:\n\n additional/Type_1/3068.jpg\n\nI try to load this with keras image preprocessor.\n\n [oystein@jupiter ]$ python\n Python 3.6.1 (default, Mar 27 2017, 00:27:06) \n [GCC 6.3.1 20170306] on linux\n Type \"help\", \"copyright\", \"credits\" or \"license\" for more information.\n >>> from keras.preprocessing import image\n Using Theano backend.\n >>> img = image.load_img(\"./additional/Type_1/3068.jpg\", target_size=(224, 224))\n Traceback (most recent call last):\n File \"", "183225": "Had same problem, but did know with which file. Now at least I know ;) \n\nValueError: output of generator should be a tuple (x, y, sample_weight) or (x, y). Found: None", "183233": "Thanks for your confirmation: Here are my corrupted(?) images:\n\n /additional/Type_1/3068.jpg\n /additional/Type_1/5893.jpg (empty)\n /additional/Type_2/2845.jpg (empty)\n /additional/Type_2/5892.jpg (empty)\n /additional/Type_2/7.jpg \n /train/Type_1/1339.jpg\n\nOf these six files, three of them is empty. That's not a mystery. The other three files are all JFIF standard 1.01, but there is also other files in this standard that loads perfectly. I guess I can open these three files in GIMP or something similar and save them in another standard. That may help.", "183546": "I haven't dug into the additional data, but I have the same problem with 1339.jpg. I was wondering if my download was corrupted, but if others are seeing the same thing there is probably a problem with the original data source.", "183598": "The MD5 hash of additional/Type_1/3068.jpg is 0B8CFE3F0D6256532DADE1C1216024D8 it's a valid file.\nThose are empty on my side too :\nadditional/Type_1/5893.jpg\nadditional/Type_2/2845.jpg\nadditional/Type_2/5892.jpg\nOthers are corrupted but I can still read them (partially) with C++ OpenCV." }, "source": "meta" }