amineHorseman / facial-expression-recognition-using-cnn

Deep facial expressions recognition using Opencv and Tensorflow. Recognizing facial expressions from images or camera stream
GNU General Public License v3.0
445 stars 141 forks source link

error while using predict.py #27

Closed 0xinit closed 4 years ago

0xinit commented 4 years ago

hi,

I'm constantly getting this error whenever I try to run predict.py. I am not sure what is wrong. Is it because of some package update?

Use standard file APIs to check for files with this prefix. /home/gaurav/anaconda3/envs/openface/lib/python2.7/site-packages/skimage/feature/_hog.py:150: skimage_deprecation: Default value of block_norm==L1 is deprecated and will be changed to L2-Hys in v0.15. To supress this message specify explicitly the normalization method. skimage_deprecation) /home/gaurav/anaconda3/envs/openface/lib/python2.7/site-packages/skimage/feature/_hog.py:248: skimage_deprecation: Argument visualise is deprecated and will be changed to visualize in v0.16 'be changed to visualize in v0.16', skimage_deprecation) Traceback (most recent call last): File "predict.py", line 93, in emotion, confidence = predict(image, model, shape_predictor) File "predict.py", line 57, in predict hog_features = sliding_hog_windows(image) File "predict.py", line 47, in sliding_hog_windows cells_per_block=(1, 1), visualise=False)) File "/home/gaurav/anaconda3/envs/openface/lib/python2.7/site-packages/skimage/feature/_hog.py", line 211, in hog g_row, g_col = _hog_channel_gradient(image) File "/home/gaurav/anaconda3/envs/openface/lib/python2.7/site-packages/skimage/feature/_hog.py", line 41, in _hog_channel_gradient g_col[:, 0] = 0 IndexError: index 0 is out of bounds for axis 1 with size 0

kaavyaastalin commented 4 years ago

I think, you are using python 2.7 version, Try to update the python version and try to run the code.

check it out below link to upgrade the python version:

https://stackoverflow.com/questions/41535881/how-do-i-upgrade-to-python-3-6-with-conda

If you are using anaconda spyder to run the rode, try to use the virtual environment. Meantime update all packages in the virtual environment. : )

0xinit commented 4 years ago

Yes, I didn't realize that python3 was required. There were some other errors, but I got them done. Thanks a lot.