Please log in to watch this conference skillscast.

Google’s Deep Dream images hit the headlines last year as the surreal and psychedelic images captured the public’s imagination. In this talk you will learn the surprisingly simple algorithm that is used to produce these images. You will discover how the “convolutional neural network” networks used to produce these images work and find out about the “model zoo” where ready trained neural networks can be found. You will explore more about the underlying structure of the neural network and how this affects the images generated. You will also learn about these models can used to classify images – a perhaps more practical application of this technology.

YOU MAY ALSO LIKE:

• Expression Oriented Programming with F# (SkillsCast recorded in April 2015)
• Keynote: The Bottom of the Iceberg: Enabling Data Science Research with F# (SkillsCast recorded in April 2019)
• Keynote: Why F# Changes Lives (SkillsCast recorded in April 2019)