/Deep-Learning-in-Computer-Vision

This repository contains material for the python workshop on deep learning in computer vision

Primary LanguageJupyter NotebookMIT LicenseMIT

Deep-Learning-in-Computer-Vision

This repository contains material for the python workshop on deep learning in computer vision.

We start with the basics of python, and get to work with some of the fundamental python libraries for data science such as: pandas, numpy, matplotlib, seaborn, scikit-learn, and more. We then dive into Keras with tensorflow backend and focus on image classification tasks. The Convolutional Neural Network (CNN) models we create varry from fully training to transfer learning to fine-tuning with transfer learning. Students will complete this workshop by working on a dataset of their choice to train a fully functioning model using Keras.

This repository can be accessed via this short link:

bit.ly/dl-in-cv

Textbooks

Acknowledgment

  • Notebooks in Python Basics folder come from A Whirlwind Tour of Python (free 100-page pdf) by Jake VanderPlas (under CC0 license). Slight modifications and updates have been made in some places to keep its content up to date. A Whirlwind Tour of Python is a fast-paced introduction to essential components of the Python language for data science and/or scientific programming. This material was written and tested using Python 3.7, and should work for any Python 3.X version.

  • Some content of a handful of notebooks come from Python Data Science Handbook by Jake VanderPlas (under the MIT license. Read more at the Open Source Initiative). Modifications and updates have been made.

  • The content of notebook 12-Exploratory-Data-Analysis.ipynb comes primarily from a Kaggle notebook by Aguiar.

  • Notebooks on the second half of the workshop are corresponding chapters of Deep Learning with Python (2nd edition)