imagenet classification task using vgg16 and alexnet in Tensoflow
#Files and Folders description
vggModel.py # defines the structure of the vgg16 model in tensorflow alexnet.py # defines the structure of the alexnet model in tensorflow caffe_classes.py # contains the 1000 classes structure related to the imagenet classification task train.py # run train.py to perform imagenet classication task using either the vgg16 or alexnet. This uses transfer learning datagenerator # helper for generating batches and rescaling images using opencv download_data # downloading the imagenet dataset from the imagenet website createtrainfiles.py # creating a file for referncing the paths of .jpeg data images on the system
#preTrainedWeights 1.bvlc_alexnet.npy [download link: http://www.cs.toronto.edu/~guerzhoy/tf_alexnet/bvlc_alexnet.npy] 2.vgg16_weights.npz [download link: https://www.cs.toronto.edu/~frossard/vgg16/vgg16_weights.npz]
#dataFiles images-class0 images-class1 images-class2
class0.txt class1.txt class2.txt
#trainData 1.Train.txt 2.valid.txt
#Run to Run just run train.py and also make the appropriate changes for the model being used and which layers to train using trasnfer learning