This totualrial demonstrates an algorithm of ultrasound image simulation using generative adversarial networks (GANs), using a data set described in a published research paper. The deep convolutional generative adversarial networks (DCGAN) is implemented in TensorFlow and PyTorch.
The following animates the randomly sampled simulations from 1st to 75th ephoch.
The local folder legacy contains an implementation of the original spatially-conditioned GANs in TensorFlow 1.