ml4a is a collection of tools and educational resources which apply techniques from machine learning to arts and creativity.
Extract a segmentation mask of an image's foreground.
Visualize and compose abstract textures using neural network feature detectors.
Generate images with the style characteristics of other images.
Generate photorealistic imagery from semantic label maps.
State-of-the-art generative modeling for images.
Image upsampling and photo enhancement.
Convert a photograph into a cartoon.
Segment the pixels of an image into their respective object categories.
Generate audio speech samples from text prompts.
Lossless inversion of face images into the latent space of a generative model.
Detect, parse, identify, and track faces in images.
Convert photographs into contour drawings or sketches.
Animate mouths in images and videos, lip-syncing from audio recordings.
Encode real images into the latent space of a pre-trained GAN
Compute optical flow between images to detect and measure motion.