This course introduces deep learning models for solving computer vision tasks in TensorFlow. It focuses on the heart of deep learning image models: convolutional neural networks.
Expert Lucas Adams shows you how to quickly get these models up to speed, especially in domains that have limited computing resources or training data. He also shows you how you can modify the architecture of a network to make it more specific to different tasks. Basic deep learning concepts such as the multilayer perceptron and linear algebra, Jupyter notebooks and the basics of building TensorFlow programs should be familiar to learners.
- Learn why the convolutional neural networks works so well for vision tasks
- Find out how each component of the architecture contributes towards prediction
- Learn how to run models using weights that have been pre-trained on large datasets and many processing hours
- Learn how to modify networks that have been trained for different tasks
- Pre-trained models can be adapted to a dataset using knowledge from an initial training run
At Jet.com, Lucas Adams is a senior-level machine learning engineer. He deploys TensorFlow to support natural language processing and computer vision. Lucas is a TensorFlow user and contributor since its release in November 2015. He holds a degree from Brown University in Applied Mathematics.
