TensorFlow, the open-source framework for deep learning created by Google, has experienced explosive growth in popularity. Its powerful syntax allows for distributed computation, improved efficiency, and modularization, enabling developers to turn high-level designs into the low-level mathematical operations required for machine learning algorithms.

In this full-day course, you will learn how to use TensorFlow for deep-learning-based computer vision applications through a combination of lectures, examples, and hands-on cloud-based laboratory exercises:

• Introduction to TensorFlow

• Graphs and Tensors

• Basics of Building Graphs in TensorFlow

• Optimization

• Linear Regression Example and Lab Exercise

• Image Classification

• Simple Neural Network to Recognize Hand-Written Digits (MNIST)

• Lab Exercise

• Deep Neural Network Models

• Multiple Layers, Nonlinearities

• Deep Neural Network with MNIST

• Lab Exercise

• Visualizing Model Operation

• TensorBoard

• Lab Exercise

• Convolutional Neural Networks (CNNs)

• Convolution, Stride, Padding, Pooling

• Lab Exercise

• Transfer Learning

• Data Sets, Data vs. Model Effort

• Transfer Learning / Retraining

• Wrap Up and Q&A

Registration Terms and Conditions:

This class is presented by the Embedded Vision Alliance, an industry association of companies developing and using computer vision technology. By registering for this class you agree to the following terms and conditions:

1. Your contact information may be used by the Alliance to inform and update you with information about activities of the Alliance and its Members and may be shared with Members of the Embedded Vision Alliance for their use in contacting you regarding their products and services.
2. During the Class, you may be photographed and/or videorecorded; the Alliance may use, reproduce, distribute, and broadcast your image, including your name, voice, likeness, and affiliation captured in such recordings as well as your name for promoting, publicizing, or explaining the Alliance and/or the Class.
3. Registration fees are non-refundable within two weeks of the class date.