Abstract:Class-conditional generative models have gained popularity due to their characteristics of learning disentangled representations. However, these models typically require labeled examples in training. In this paper, we explore the feasibility of training these models on completely unlabeled data, under the assumption that we have access to other labeled data. The labeled data share the same label space, while their domain is shifted. Our model, which we refer to as KTransGAN, incorporates a classifier to transfer knowledge from the labeled data and performs collaborative learning with the conditional generator. By adopting these measures, KTransGAN is able to approximate the conditional distribution of the unlabeled data and simultaneously introduces a new solution to the unsupervised domain adaptation problem. To mitigate the training difficulty of our generative adversarial networks-based model, variational encoding and feature matching are also considered. From the empirical results, KTransGAN exhibits outstanding performance on a number of synthetic datasets and multiple real-world benchmarks. The quality of the synthesized instances is far superior to the pure variational autoencoding model. For example, on the CIFAR-10 dataset, our model scores 35.3 in FID, while the other model scores 128.45. In addition, the synthesis quality is close to the case when the model is trained in a fully supervised setting over the same number of training iterations. Regarding the classification performance, for instance, our model surpasses the highest state-of-the-art results (89.19%) by a large margin and achieves a test accuracy of 95.31% on the unlabeled data SVHN, while MNIST represents the labeled data. These results highlight the effectiveness of our proposed framework.

Enhancing Triplegan For Semi-Supervised Conditional Instance Synthesis And Classification

Statistics Enhancement Generative Adversarial Networks for Diverse Conditional Image Synthesis

Dual Distribution Matching GAN

Triple Generative Adversarial Networks.

Semi-Supervised Generative Learning with Extended Distribution Matching for Class-Conditional Image Synthesis

Semantic Regularized Class-Conditional GANs for Semi-Supervised Fine-Grained Image Synthesis

Regularizing Discriminative Capability Of Cgans For Semi-Supervised Generative Learning

Semi-Supervised Single-Stage Controllable GANs for Conditional Fine-Grained Image Generation

Enhanced self-supervised GANs with blend ratio classification

Enhancing Stability in Training Conditional Generative Adversarial Networks via Selective Data Matching

Improving GANs with A Dynamic Discriminator

Improving GAN Training via Feature Space Shrinkage

Distribution Enhancement for Imbalanced Data with Generative Adversarial Network

Image Augmentations for GAN Training

Conditional Face Synthesis for Data Augmentation.

Semi-Supervised Learning with Flexible Discriminator Objective in Generative Adversarial Networks Framework

Transformation GAN for Unsupervised Image Synthesis and Representation Learning

Student's T-Generative Adversarial Networks

MatchGAN: A Self-Supervised Semi-Supervised Conditional Generative Adversarial Network

KTransGAN: Variational Inference-Based Knowledge Transfer for Unsupervised Conditional Generative Learning

TripletGAN: Training Generative Model with Triplet Loss.