crypdick/awesome-neural-art

A curated list of awesome neural network-based art resources.

Awesome Neural Art

Creating art and manipulating images using deep neural networks.

Contents

• Colorization
• Style Transfer
• Super Resolution
• Patching
• Background subtraction
• Deep Dream
• Interactive art
• Multiple categories
• Lectures

Colorization

See also: Awesome Image Colorization and the colorization subreddit where people do this manually.

• Interactive Deep Colorization

• colornet

• Colorful Image Colorization

Style Transfer

• neural-style - Torch implementation of neural style algorithm.

• adaptive-style-transfer - Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization.
• deep-photo-styletransfer - Code and data for "Deep Photo Style Transfer".
• fast-style-transfer - TensorFlow CNN for fast style transfer with larger scale style features in transformations.
• texture_nets - Code for "Texture Networks: Feed-forward Synthesis of Textures and Stylized Images".
• AdaIN-style - Code for Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization. Contains neat features such as transfering style without color, style interpolation, and spatial control.

• Bicycle GAN - NIPS 2017 Toward Multimodal Image-to-Image Translation, PyTorch implementation. Given the same night image, model is able to synthesize possible day images with different types of lighting, sky and clouds.

• Neural Style Transfer - Keras Implementation of Neural Style Transfer from the paper "A Neural Algorithm of Artistic Style" in Keras 2.0+.
• Deep Painterly Harmonization - Photoshopping an object into a painting, and then neural network changes the style (pallete, strokes, luminosity, etc) of the object to match the painting style. See paper.

Super Resolution

• waifu2x - Image Super-Resolution for Anime-Style Art.

• srgan - Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network.

• fft-descreen - Clean up deconvolution checkerboard artifacts found in style transferred images.

Patching

AKA in-painting

• DeepFill V1 - Generative Image Inpainting with Contextual Attention.
• EdgeConnect - a two-stage adversarial model that comprises of an edge generator followed by an image completion network.

- [Deep Image Completion](https://github.com/adamstseng/general-deep-image-completion) - Contains face-completion and general image completion models. Each can complete images with differnet types of corrupted masks like scribbles, lines, dots and texts.

Background subtraction

AKA image matting. See also: Awesome Background Subtraction

• Deep Image Matting - Implementations of the Deep Image Matting paper.

  • Dataset generator - Generates training data for the Deep Image Matting paper.
  • Tensorflow implementation - Original paper author.
  • PyTorch implementation - Same author as Keras implementation, but this is their improved (and maintained) codebase.
  • Another PyTorch implementation

• Salient Object Detection - Tensorflow implementation for cvpr2017 paper "Deeply Supervised Salient Object Detection with Short Connections" (not peer reviewed). Same author as original Deep Image Matting author.

• AlphaGAN - Unofficial implementation of AlphaGAN: Generative adversarial networks for natural image matting. Uses a GAN as the name suggests.

• DeepLab V3+ - Implementations of an encoder-decoder Modified Aligned Xception and ResNet as backbone (paper).

Deep Dream

• deepdream - The original Google implementation.

Interactive art

• neural-doodle - Turn your two-bit doodles into fine artworks with deep neural networks, generate seamless textures from photos, transfer style from one image to another, perform example-based upscaling, but wait... there's more! (An implementation of Semantic Style Transfer.)

• iGAN - iGAN: Interactive Image Generation via Generative Adversarial Networks.

• style2paints - An AI-driven interactive line art colorization tool. Backend is a Residual U-net and Auxiliary Classifier GAN.

Multiple categories

• deep-image-prior - Image restoration with neural networks but without learning. Does artifact removal, inpainting, super-resolution, denoising.

Lectures

• The Neural Aesthetic @ ITP-NYU, Fall 2018
• Neural Style Transfer Review - Repo for the Neural Style Transfer review article.

License

To the extent possible under law, Richard Decal has waived all copyright and related or neighboring rights to this work.