This course provides an introduction to using deep learning models to solve computer vision tasks in TensorFlow and is focused on the work horse of deep learning image models: the convolutional neural network.
Expert Lucas Adams teaches you how to get these models up and running fast, especially in domains with limited computing resources or training data, and shows you how to modify the architecture of a neural network to make the model specialized to different tasks. Learners should be familiar with basic deep learning concepts like the multilayer perceptron, linear algebra, Jupyter notebooks, and the basics of building and running TensorFlow programs.
- Understand why the convolutional neural network works so well for vision tasks
- Explore how each component of the architecture contributes to prediction
- Learn to run models using weights pre-trained on large datasets using many processing hours
- Discover how to modify pre-trained networks for completely different tasks
- Tune pre-trained models to a dataset while using knowledge stored from an initial training run
Lucas Adams is a senior level machine learning engineer at Jet.com, where he deploys TensorFlow for computer vision and natural language processing systems. A user and contributor to TensorFlow since its release in November 2015, Lucas holds a degree in Applied Mathematics from Brown University.