Machine Learning for Art

ml4a is a collection of tools and educational resources which apply techniques from machine learning to arts and creativity.

Models Fundamentals ml5.js

Salient object detection
Salient object detection

Extract a segmentation mask of an image's foreground.

  View code   Open in Colab
DeepDream
DeepDream

Visualize and compose abstract textures using neural network feature detectors.

  View code   Open in Colab
Neural style transfer
Neural style transfer

Generate images with the style characteristics of other images.

  View code   Open in Colab
Image-to-image translation
Image-to-image translation

Generate photorealistic imagery from semantic label maps.

  View code   Open in Colab
StyleGAN2

State-of-the-art generative modeling for images.

  View code   Open in Colab
Super-resolution
Super-resolution

Image upsampling and photo enhancement.

  View code   Open in Colab
Cartoonization
Cartoonization

Convert a photograph into a cartoon.

  View code   Open in Colab
Semantic segmentation
Semantic segmentation

Segment the pixels of an image into their respective object categories.

  View code   Open in Colab
Text-to-speech synthesis
Text-to-speech synthesis

Generate audio speech samples from text prompts.

  View code   Open in Colab
Reversible generative models
Reversible generative models

Lossless inversion of face images into the latent space of a generative model.

  View code   Open in Colab
Processing faces
Processing faces

Detect, parse, identify, and track faces in images.

  View code   Open in Colab
Photo sketching
Photo sketching

Convert photographs into contour drawings or sketches.

  View code   Open in Colab
Lip-syncing videos
Lip-syncing videos

Animate mouths in images and videos, lip-syncing from audio recordings.

  View code   Open in Colab
GAN inversion
GAN inversion

Encode real images into the latent space of a pre-trained GAN

  View code   Open in Colab
Optical flow
Optical flow

Compute optical flow between images to detect and measure motion.

  View code   Open in Colab