Closed klvn930815 closed 7 years ago
Hi!
Thank you for your comment, you are completely right, I will change it on the code and upload it again! Sorry for the inconveniences,
Míriam
Thanks for your quick reply! And I have another question. Since the pretrained VGG16 model was trained with BGR images, so I need to convert the training images into BGR order before using the pretrained VGG16 model in my own task, is it right? Thanks.
Yes! The order should be BGR, and the mean of Imagenet database for each channel should be subtracted.
OK, Thank you very much!
Hello, all. I found that the code used the pretrained VGG16 model from https://gist.github.com/baraldilorenzo/07d7802847aaad0a35d3,and because of using opencv, the input was loaded in BGR order. But I found that in keras.preprocessing, the image was load in RGB order , https://github.com/fchollet/keras/blob/master/keras/preprocessing/image.py#L300 was there something wrong? And my environment is Keras 1.2.1 ,Theano 0.8.2 , Tensorflow 0.12.1 , using Tensorflow backend.
Thanks!