Bookmark and Share

Upcoming Training Classes Cover Deep Learning for Vision Using TensorFlow

On July 13, 2017 at the Hyatt Regency Hotel in Santa Clara, California, the Embedded Vision Alliance will deliver a full-day hands-on training class on Google's TensorFlow deep learning framework. Here's the description, from the event page:

This full-day course will provide you with the hands-on knowledge you need to develop deep learning computer vision applications—both on embedded systems and in the cloud—with TensorFlow, one of today's most popular frameworks for deep learning.

If you’re developing deep learning-based computer vision applications and want to use TensorFlow to do it, this course is for you! The only pre-requisites are a good understanding of Python (e.g., basic syntax and structure) and a basic knowledge of deep learning concepts.

Course Outline

Morning:

  • Introduction to TensorFlow
    • What It Is, What It’s Used For
    • Graphs and Tensors
    • Basics of Building Graphs in TensorFlow
  • Neural Networks in TensorFlow
  • TensorBoard Visualization Tool
  • Exercises including Fibonacci Sequence, Linear Regression and Recognition of Hand-Written Digits (MNIST Database)

Afternoon:

  • Object Recognition in TensorFlow
    • Quick Refresher on Convolutional Neural Networks (CNNs)
    • Exploration of CNN model for MNIST demo
    • Training Data and Issues
  • Open Source CNN Models
  • Writing your own CNN model: TF-Slim and Keras
  • Exercises including  CNN Model for MNIST, Transfer Learning for Object Recognition, Object recognition from video

The price is $895. For more information and to register, see the event page.

The training class will also be held in Hamburg, Germany on September 7, 2017. You can also sign up on the event page to receive updates on future planned classes.