Abstract:Semi-supervised learning, which entails training a model with manually labeled images and pseudo-labels for unlabeled images, has garnered considerable attention for its potential to improve image classification performance. Nevertheless, incorrect decision boundaries of classifiers and wrong pseudo-labels for beneficial unlabeled images below the confidence threshold increase the generalization error in semi-supervised learning. This study proposes a novel framework for semi-supervised learning termed consistency-regularized bad generative adversarial network (CRBSGAN) through a new loss function. The proposed model comprises a discriminator, a bad generator, and a classifier that employs data augmentation and consistency regularization. Local augmentation is created to compensate for data scarcity and boost bad generators. Moreover, label consistency regularization is considered for bad fake images, real labeled images, unlabeled images, and latent space for the discriminator and bad generator. In the adversarial game between the discriminator and the bad generator, feature space is better captured under these conditions. Furthermore, local consistency regularization for good-augmented images applied to the classifier strengthens the bad generator in the generator–classifier adversarial game. The consistency-regularized bad generator produces informative fake images similar to the support vectors located near the correct classification boundary. In addition, the pseudo-label error is reduced for low-confidence unlabeled images used in training. The proposed method reduces the state-of-the-art error rate from 6.44 to 4.02 on CIFAR-10, 2.06 to 1.56 on MNIST, and 6.07 to 3.26 on SVHN using 4000, 3000, and 500 labeled training images, respectively. Furthermore, it achieves a reduction in the error rate on the CINIC-10 dataset from 19.38 to 15.32 and on the STL-10 dataset from 27 to 16.34 when utilizing 1000 and 500 labeled images per class, respectively. Experimental results and visual synthesis indicate that the CRBSGAN algorithm is more efficient than the methods proposed in previous works. The source code is available at https://github.com/ms-iraji/CRBSGAN ↗.

CCS-GAN: a semi-supervised generative adversarial network for image classification

Image classification with consistency-regularized bad semi-supervised generative adversarial networks: a visual data analysis and synthesis

Self-supervised Graph Representations with Generative Adversarial Learning

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

Generative-Discriminative Complementary Learning

Deep Graph-Convolutional Generative Adversarial Network for Semi-Supervised Learning on Graphs

CDGAN-BERT: Adversarial constraint and diversity discriminator for semi-supervised text classification

GAN-CL: Generative Adversarial Networks for Learning from Complementary Labels

Good Semi-supervised Learning That Requires a Bad GAN.

Relaxed Conditional Image Transfer for Semi-supervised Domain Adaptation

Consistency Regularization with Generative Adversarial Networks for Semi-Supervised Learning

Robust Semi-Supervised Classification using GANs with Self-Organizing Maps

Semi-supervised community detection method based on generative adversarial networks

Weakly-supervised Generative Adversarial Networks for medical image classification

Semi-supervised semantic image segmentation using dual discriminator adversarial networks

Enhanced self-supervised GANs with blend ratio classification

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

CDL-GAN: Contrastive Distance Learning Generative Adversarial Network for Image Generation

A Supervised Progressive Growing Generative Adversarial Network for Remote Sensing Image Scene Classification

GL-GAN: Adaptive Global and Local Bilevel Optimization model of Image Generation

Generative Adversarial Network for Semi-Supervised Image Captioning