MGAN: Training Generative Adversarial Nets with Multiple Generators

By Quan Hoang, Tu Dinh Nguyen, Trung Le, Dinh Q. Phung. 2018

250 citations recorded by the source corpus.

Subfields: Generative Models, Optimization & Training

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Generative Models

In Optimization & Training

Citations in the corpus

Dual Contradistinctive Generative Autoencoder (2020, neutral): “1 MGAN [27] 8.”
A Comprehensive Survey of Neural Architecture Search (2020, background): “Compared with the most advanced manual GANs [94–97], AutoGAN is very competitive.”
AutoGAN: Neural Architecture Search for Generative Adversarial Networks (2019, background): “In terms of inception score, AutoGAN is slightly next to Progressive GAN [26], and surpasses many latest strong competitors such as SN-GAN [42], improving MMD-GAN [56], Dist-GAN [54], MGAN [19], and WGAN-GP [13].”
Flow Matching for Generative Modeling (2022, neutral): “BigGAN (Luˇci´cetal.,2019) – 25.3 PGMGAN (Armandpouretal.,2021) – 21.7 FM w / OT 2.90 20.9 2017)), and averaged number of function evaluations (NFE) required for the adaptive solver to reach its a prespecified numerical tolerance, averaged over 50k samples.”
A Disentangling Invertible Interpretation Network for Explaining Latent Representations (2020, background): “…to reconstruct images from a representation whose marginal distribution is matched to a standard normal distribution, or on Generative Adversarial Networks [11, 14, 39], which directly map samples from a standard normal distribution to realistic looking images as judged by a discriminator network.”
Evolutionary Generative Adversarial Networks (2018, background): “Mixture GAN [50] is proposed to overcome the mode collapse problem by training multiple generators to specialize in different data modes.”
An Improved Evaluation Framework for Generative Adversarial Networks (2018, background): “Researchers have been continuously developing better GAN architectures [18,19] and training strategies [20,21] on generating domainspecific images.”
PassGAN: A Deep Learning Approach for Password Guessing (2017, unknown): “[25], there have been several improvements on GANs [3, 4, 6–8, 10, 14, 16, 27, 35, 36, 42, 47, 53, 55, 56, 58–60, 63, 66, 69, 74, 76, 80, 87, 90, 94, 96], where each new paper provides novel improvements in the domain.”
All in One Bad Weather Removal Using Architectural Search (2020, background): “Third, most existing discriminators in GAN-based approaches are trained to judge whether the restored images are real or not.”
TransGAN: Two Transformers Can Make One Strong GAN (2021, background): “…the strong Auto-GAN (Gong et al., 2019) and many other latest competitors such as SN-GAN (Miyato et al., 2018), improving MMD-GAN (Wang et al., 2018a), and MGAN (Hoang et al., 2018), in terms of inception score (IS); it is only next to the huge and highly-engineered Progressive GAN and StyleGAN v2.”
Hausdorff GAN: Improving GAN Generation Quality With Hausdorff Metric (2021, methodology): “We take typical multigenerator GAN variants, which are MGAN [23], disconnected manifold learning GAN with prior learning (DMGAN-PL) [27], and localized GAN (LGAN) [36], respectively, as the example to illustrate our analysis.”
Generative Adversarial Network Training is a Continual Learning Problem (2018, methodology): “There have been several proposed models utilizing multiple generators (Hoang et al., 2018; Ghosh et al., 2018) or multiple discriminators (Durugkar et al., 2017; Neyshabur et al., 2017), while Bayesian GAN (Saatchi & Wilson, 2017) considered distributions on the parameters of both networks, but…”
ProbGAN: Towards Probabilistic GAN with Theoretical Guarantees (2018, methodology): “, 2017) and the recent state-of-art method, MGAN (Hoang et al., 2018).”
Instance-Conditioned GAN (2021, unknown): “While these methods are promising and yield visually appealing samples, their quality is still far from those obtained with class-conditional GANs.”
Partition-Guided GANs (2021, background): “As shown in Table 2, PGMGAN achieves state-ofthe-art FID and IS on the STL-10 dataset.”