Repurposing Existing Deep Learning Models: Image Similarity for fun (& profit?)

In this presentation I will walk through the story of spending an afternoon exploiting an existing trained deep learning model, extracting deep features, and transferring them to the task of building an image similarity service for house pictures. I will show how you don't need to be an expert in deep learning or convolutional neural networks in order to leverage deep features in your next application. The presentation will be entirely delivered from an IPython Notebook and written in Python, using Dato for extracting deep features from a trained deep learning model and deploying the image similarity service.