How Google’s Deep Dream Works

The following video is a general description of how Google’s Deep Dream computer vision (image recognition) works.  Deep Dream uses a convolutional neural network to look for patterns in an image or video based on large sets of images previously analyzed for defining characteristics.  The beauty of Deep Dream is that it can modify images to accent fragments within the image to more closely resemble the aspects from images it has previously evaluated.  The effects of such augmentation can be quite striking; creating strange, almost psychedelic, results.

Runtime: 13:42

This video can also be found here.

Video Info:

Published on Aug 26, 2016

Surreal images created by Google’s Deep Dream code flooded the internet in 2015 but how does deep dream do it? Image analyst Dr Mike Pound.

Inside a Neural Network:
Cookie Stealing:
FPS & Digital Video:
Password Cracking:
Gamer’s Paradise:

Images seen/manipulated in this video:…

This video was filmed and edited by Sean Riley.

Computer Science at the University of Nottingham:

Computerphile is a sister project to Brady Haran’s Numberphile. More at

  • Category

  • License

    • Standard YouTube License

One thought on “How Google’s Deep Dream Works

  1. Tobiah Townsend February 4, 2018 / 7:33 pm

    This reminds me of the old “psychedelic screen saver”… (95?)


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s