/keras-transfer-learning

Transfer learning of ImageNet models using Python and Keras

Primary LanguageJupyter Notebook

Transfer Learning

Simple examples for pre-trained Keras deep learning models on images based on this blogpost.

Base model used is the VGG16 model.

Datasets

Datasets need to be stored in a subfolder in ./data/ where images belonging to different classes go in separate subfolders. For instance: ./data/inria/pos/ for the INRIA pedestrian dataset people images and ./data/inria/neg/ for negative class images.