Transfer Learning

The intuition behind transfer learning for image classification is that if a model is trained on a large and general enough dataset, this model will effectively serve as a generic model of the visual world. You can then take advantage of these learned feature maps without having to start from scratch by training a large model on a large dataset. A pre-trained model is a saved network that was previously trained on a large dataset, typically on a large-scale image-classification task. You either use the pre-trained model as is or use transfer learning to customize this model to a given task.

watch a couple of video tutorials on why case studies helps in better classification and look at some of the standard pre-trained network architectures.

Try to read this paper on AlexNet, other that the complex mathematical notation, you should be able to map the various concepts you have learnt and how they are configured for the ImageNet competition.

Now, with the understanding of transfer learning fundamental concepts, for the rest of the day you will be researching how to implement transfer learning techniques, followed by implementing them to your creative brief project.

As usual, tensorflow has the stored models for us to use. Follow this tensorflow tutorial or tensorflow guide as your primary research source and use mobilenet v2 as your base model to train your creative brief classifier as well.