Figure 2.

A high-level diagram of our method. Users input a set of transformation functions h₁, …, h_K and unlabeled data. A generative adversarial approach is then used to train a null class discriminator, D^∅, and a generator, G, which produces TF sequences $h_{{\tau }_1},\dots ,h_{{\tau }_L}$. Finally, the trained generator is used to perform data augmentation for an end discriminative model D^f.