Neural-dream

Neural-Dream: Open Source Deep Neural Network Software

Open source software generating artistic imagery with deep neural networks, enhancing patterns and textures in existing images to create dreamlike, psychedelic versions.

What is Neural-dream?

neural-dream is an open source software project that utilizes deep neural networks to generate new images. It was created in 2015 by Alexander Mordvintsev from Google. The software is based on convolutional neural networks which are trained on a large image dataset.

When you input an image into neural-dream, it detects the visual patterns such as shapes, textures, and colors. It then enhances and amplifies these patterns in the output image to create a hallucinogenic, dreamlike version. The results are abstract, psychedelic images that look like something you might see in a dream or vision.

Some key features and capabilities of neural-dream include:

• Works with pre-trained convolutional neural network models or allows you to train your own model
• Multiple algorithms for enhancing and mutating visual features in images
• Control parameters to adjust the diversity and intensity of the neural patterns
• Batch processing capability to generate neural dream imagery from multiple source images
• Flexible framework that has been ported to multiple platforms and programming languages

The software is popular among digital artists and creators looking to generate new kinds of algorithmic artwork. The images have an abstract quality since shapes and objects from the original image get distorted and blended together in imaginative ways.

Neural-dream Features

Official Links