During the E2-Create project, several libraries and example code have been developed for two popular programming frameworks: Pytorch and openFrameworks. PyTorch is a popular open source framework for developing, training, and deploying deep learning models. openFrameworks is an open source C++ programming framework for creative coding.
The articles in this section offer a brief introduction for these two frameworks and document several coding examples. In the case of PyTorch, the examples involve machine learning models for generating synthetic images, dance poses, and dance movements. In the case of openFrameworks, the examples involve addons for creating a processing pipeline for motion capture and for implementing simulations of mass-spring systems, articulated morphologies, and groups of flocking entities.
The following articles are currently available:
- Pytorch
- Image Data
- Motion Capture Data
- Utility Classes and Functions
- Pose Sequence Generation (RNN)
- Pose Sequence Generation (RNN+MDN)
- Image Generation with a GAN
- Pose Generation with a GAN
- Pose Sequence Generation with a GAN
- Image Generation with an Adversarial Autoencoder
- Pose Generation with an Adversarial Autoencoder
- Pose Sequence Generation with an Adversarial Autoencoder
- Granular Dance
- openFrameworks