search

0x5eba / Skin-Cancer-Segmentation

Classification and Segmentation with Mask-RCNN of Skin Cancer using ISIC dataset
MIT License
80 stars 24 forks source link

main.py not working #6

Closed Asmar-git closed 5 years ago

Asmar-git commented 5 years ago

Yes im sure the DATA folder is not empty

Asmar-git commented 5 years ago

when i execute python3 main.py and add the data path it gives the following errors: hali@mivl-1:~/Skin-Cancer-Segmentation$ python3 main.py Using TensorFlow backend. Insert the path of Data [ Link /home/../ISIC-Archive-Downloader/Data/ ] : /home/hali/ISIC-Archive-Downloader/Data/ Traceback (most recent call last): File "main.py", line 139, in image_ids = np.random.choice(dataset_train.image_ids, 3) File "mtrand.pyx", line 1126, in mtrand.RandomState.choice ValueError: a must be non-empty

0x5eba commented 5 years ago

You have to give me more details.

1) First make sure that the folder is Segmentation and not Segmentations. 2) Check with both python3.7 and python3.6 using the requirement.txt in the repo 3) Print dataset_train (only the first 5 results) 4) Print dataset_train.image_ids (only the first 5 results)

Asmar-git commented 5 years ago

Insert the path of Data [ Link /home/../ISIC-Archive-Downloader/Data/ ] : /home/hali/ISIC-Archive-Downloader/Data/ Traceback (most recent call last): File "main.py", line 139, in image_ids = np.random.choice(dataset_train.image_ids, 3) File "mtrand.pyx", line 773, in numpy.random.mtrand.RandomState.choice ValueError: 'a' cannot be empty unless no samples are taken This Occurs with python3.7 the folder name is Segmentation

Asmar-git commented 5 years ago

What i think the issue is that while loading the images you think of them as jpg files, but what i have is jpeg images

0x5eba commented 5 years ago

Change https://github.com/0x5eba/Skin-Cancer-Segmentation/blob/master/main.py#L108 .jpg to .jpeg. Also, as I said, follow the instruction that I used in the readme, and should work.

Asmar-git commented 5 years ago

i've followed all the instructions, changed jpg to jpeg in line 108, still not working

On Thu, 19 Sep 2019 at 17:07, Sebastien Biollo notifications@github.com wrote:

Closed #6 https://github.com/0x5eba/Skin-Cancer-Segmentation/issues/6.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/0x5eba/Skin-Cancer-Segmentation/issues/6?email_source=notifications&email_token=AMQZSWXQMLR7VKQ72VPBPU3QKNTRZA5CNFSM4IXS6EG2YY3PNVWWK3TUL52HS4DFWZEXG43VMVCXMZLOORHG65DJMZUWGYLUNFXW5KTDN5WW2ZLOORPWSZGOTXA637I#event-2646732285, or mute the thread https://github.com/notifications/unsubscribe-auth/AMQZSWWOVDGXEK2YEZFWUSLQKNTRZANCNFSM4IXS6EGQ .

Asmar-git commented 5 years ago

it says : Using TensorFlow backend. Traceback (most recent call last): File "main.py", line 109, in if not img: ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

On Thu, 19 Sep 2019 at 17:40, Asmar Asim asmarasim96@gmail.com wrote:

i've followed all the instructions, changed jpg to jpeg in line 108, still not working

On Thu, 19 Sep 2019 at 17:07, Sebastien Biollo notifications@github.com wrote:

Closed #6 https://github.com/0x5eba/Skin-Cancer-Segmentation/issues/6.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/0x5eba/Skin-Cancer-Segmentation/issues/6?email_source=notifications&email_token=AMQZSWXQMLR7VKQ72VPBPU3QKNTRZA5CNFSM4IXS6EG2YY3PNVWWK3TUL52HS4DFWZEXG43VMVCXMZLOORHG65DJMZUWGYLUNFXW5KTDN5WW2ZLOORPWSZGOTXA637I#event-2646732285, or mute the thread https://github.com/notifications/unsubscribe-auth/AMQZSWWOVDGXEK2YEZFWUSLQKNTRZANCNFSM4IXS6EGQ .