Abstract:In this paper, we propose the Soft Generative Adversarial Network (SoftGAN), a strategy that utilizes a dynamic borderline softening mechanism to train Generative Adversarial Networks. This mechanism aims to solve the mode collapse problem and enhance the training stability of the generated outputs. Within the SoftGAN, the objective of the discriminator is to learn a fuzzy concept of real data with a soft borderline between real and generated data. This objective is achieved by balancing the principles of maximum concept coverage and maximum expected entropy of fuzzy concepts. During the early training stage of the SoftGAN, the principle of maximum expected entropy of fuzzy concepts guides the learning process due to the significant divergence between the generated and real data. However, in the final stage of training, the principle of maximum concept coverage dominates as the divergence between the two distributions decreases. The dynamic borderline softening mechanism of the SoftGAN can be likened to a student (the generator) striving to create realistic images, with the tutor (the discriminator) dynamically guiding the student towards the right direction and motivating effective learning. The tutor gives appropriate encouragement or requirements according to abilities of the student at different stages, so as to promote the student to improve themselves better. Our approach offers both theoretical and practical benefits for improving GAN training. We empirically demonstrate the superiority of our SoftGAN approach in addressing mode collapse issues and generating high-quality outputs compared to existing approaches.

Distribution constraining for combating mode collapse in generative adversarial networks

Distribution Fitting for Combating Mode Collapse in Generative Adversarial Networks

Distribution Fitting for Combating Mode Collapse in GANs

Soft Generative Adversarial Network: Combating Mode Collapse in Generative Adversarial Network Training Via Dynamic Borderline Softening Mechanism

Dual Distribution Matching GAN

Combating Mode Collapse via Offline Manifold Entropy Estimation

Antimode collapse generative adversarial networks

Statistics Enhancement Generative Adversarial Networks for Diverse Conditional Image Synthesis

DynGAN: Solving Mode Collapse in GANs with Dynamic Clustering

MGGAN: Solving Mode Collapse using Manifold Guided Training

Mode Seeking Generative Adversarial Networks for Diverse Image Synthesis

Combating Mode Collapse in GANs via Manifold Entropy Estimation

Reducing Mode Collapse With Monge–Kantorovich Optimal Transport for Generative Adversarial Networks

Escaping from Collapsing Modes in a Constrained Space

Study of Prevention of Mode Collapse in Generative Adversarial Network (GAN)

InvertGAN: Reducing Mode Collapse with Multi-Dimensional Gaussian Inversion

STM-GAN: Sequentially Trained Multiple Generators for Mitigating Mode Collapse

No Modes Left Behind: Capturing the Data Distribution Effectively Using GANs

On Mode Collapse in Generative Adversarial Networks