MTCNN-Tensorflow icon indicating copy to clipboard operation
MTCNN-Tensorflow copied to clipboard

shape error after some executions

Open alirazaee opened this issue 5 years ago • 2 comments

Getting this error ! . . . . . . . . . . . . . . 581 images done, pos: 9404 part: 25731 neg: 29047 581 images done, pos: 9404 part: 25731 neg: 29048 581 images done, pos: 9404 part: 25731 neg: 29049 581 images done, pos: 9404 part: 25731 neg: 29050 581 images done, pos: 9412 part: 25743 neg: 29050 581 images done, pos: 9412 part: 25751 neg: 29050 581 images done, pos: 9419 part: 25764 neg: 29050 581 images done, pos: 9424 part: 25779 neg: 29050 581 images done, pos: 9424 part: 25789 neg: 29050 Traceback (most recent call last):

File "", line 1, in runfile('E:/MTCNN-Tensorflow-master/prepare_data')

File "C:\Users\alira\Anaconda3\lib\site-packages\spyder_kernels\customize\spydercustomize.py", line 786, in runfile execfile(filename, namespace)

File "C:\Users\alira\Anaconda3\lib\site-packages\spyder_kernels\customize\spydercustomize.py", line 110, in execfile exec(compile(f.read(), filename, 'exec'), namespace)

MTCNN-Tensorflow-master/MTCNN-Tensorflow-master/prepare_data/gen_shuffle_data.py", line 163, in main()

MTCNN-Tensorflow-master/MTCNN-Tensorflow-master/prepare_data/gen_shuffle_data.py", line 79, in main height, width, channel = img.shape

AttributeError: 'NoneType' object has no attribute 'shape'

alirazaee avatar Aug 27 '19 16:08 alirazaee

yes, i have the same problem too :(( T^T please reply who already done the testing @wangbm

nyingnying avatar Sep 27 '19 10:09 nyingnying

Haiii, i already done training size 12. also got this error before : AttributeError: 'NoneType' object has no attribute 'shape'

It solved by check your wider face dataset, it need to be downloaded first. And if it is already downloaded, please check all those images need to be completed based on image list on 'wider_face_train.txt'

I also try using only one folder to make training more faster, and deleted all image list from folder 1 - 10.

please try, thanks

fmoulita avatar Oct 12 '19 07:10 fmoulita