Issue with train.py about facial-expression-recognition-using-cnn HOT 2 CLOSED

Figured out the above issue, and it was due to the fact that data for validation was empty. In convert_fer2013_to_images_and_landmarks.py file, line 169, the counter for 'nb_images_per_label' reaches IMAGES_PER_LABEL for training itself and hence when extracting Validation labels, the control never goes into 'If Loop'. Fixed this locally to increment the IMAGES_PER_LABEL counter to accommodate the # of labels we need for Training and Testing. Will create a pull request soon with few additional minor changes to accommodate when training for less than 7 expression. Otherwise, @amineHorseman, this is great piece of work from you.

from facial-expression-recognition-using-cnn.

Thank you @sandeepalur
Sorry for late reply
Yes feel free to send a pull request when you finish your modifications

from facial-expression-recognition-using-cnn.