classification - Original Image FileName in Caffe Network -
i using cifar(https://github.com/bvlc/caffe/tree/master/examples/cifar10) model in caffe. able train model , probability of wrongly classified images using pycaffe interface.
is possible name of image filename data layer (created lmdb list of image filename , label train.txt , val.txt - https://github.com/bvlc/caffe/blob/master/examples/imagenet/create_imagenet.sh)?
if not, how check content (original image's filename written in train.txt) of wrongly classified image (not labels)?
Comments
Post a Comment