On Adversarial Mixup Resynthesis

Abstract

In this paper, we explore new approaches to combining information encoded within the learned representations of auto-encoders. We explore models that are capable of combining the attributes of multiple inputs such that a resynthesised output is trained to fool an adversarial discriminator for real versus synthesised data. Furthermore, we explore the use of such an architecture in the context of semi-supervised learning, where we learn a mixing function whose objective is to produce interpolations of hidden states, or masked combinations of latent representations that are consistent with a conditioned class label. We show quantitative and qualitative evidence that such a formulation is an interesting avenue of research.

'Camera-ready draft'

Cited by in corpus (11)

• i-Mix: A Domain-Agnostic Strategy for Contrastive Representation Learning
• How Does Mixup Help With Robustness and Generalization?
• InstaHide: Instance-hiding Schemes for Private Distributed Learning
• Interpolation-based semi-supervised learning for object detection
• Autoencoder Image Interpolation by Shaping the Latent Space
• It Takes Two to Tango: Mixup for Deep Metric Learning
• Learning Flat Latent Manifolds with VAEs
• Inclusive GAN: Improving Data and Minority Coverage in Generative Models
• Augmentation-Interpolative AutoEncoders for Unsupervised Few-Shot Image Generation
• Uniform Priors for Data-Efficient Transfer
• Adversarial Mixing Policy for Relaxing Locally Linear Constraints in Mixup