DeepDream Fractal Zoom
DeepDream Fractal Zoom
This video was generated using a neural network trained for many hours, combining Google's DeepDream technology with fractal zoom imagery.
How It Was Made
The input for the neural network was the first 1500 255x255 snapshots at 1 second intervals from a fractal zoom video by "Maths Town":
https://www.youtube.com/watch?v=pCpLWbHVNhk
The video was generated via a few tweaks to Google's DeepDream project:
https://github.com/google/deepdream
About DeepDream
DeepDream is a computer vision program created by Google that uses a convolutional neural network to find and enhance patterns in images. The algorithm amplifies the patterns it detects, creating surreal, dream-like imagery.
Technical Details
- Training time: Many hours of neural network training
- Input resolution: 255x255 pixel snapshots
- Frame count: 1500 frames
- Source material: Fractal zoom animation
The combination of mathematical fractals with neural network pattern recognition creates a unique visual experience that blends mathematical precision with AI-generated psychedelic imagery.
More details on the technical implementation are planned for future content.