A Novel Framework for Selection of GANs for an Application
arXiv:2002.08641
Abstract
Generative Adversarial Network (GAN) is a current focal point of research. The body of knowledge is fragmented, leading to a trial-error method while selecting an appropriate GAN for a given scenario. We provide a comprehensive summary of the evolution of GANs starting from its inception addressing issues like mode collapse, vanishing gradient, unstable training and non-convergence. We also provide a comparison of various GANs from the application point of view, its behaviour and implementation details. We propose a novel framework to identify candidate GANs for a specific use case based on architecture, loss, regularization and divergence. We also discuss application of the framework using an example, and we demonstrate a significant reduction in search space. This efficient way to determine potential GANs lowers unit economics of AI development for organizations.
11 pages, 1 figure, 8 tables
References in corpus (16)
- GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
- Energy-based Generative Adversarial Network
- Layer Normalization
- Real-valued (Medical) Time Series Generation with Recurrent Conditional GANs
- A Review on Generative Adversarial Networks: Algorithms, Theory, and Applications
- Spectral Norm Regularization for Improving the Generalizability of Deep Learning
- Learning in Implicit Generative Models
- MAD-GAN: Multivariate Anomaly Detection for Time Series Data with Generative Adversarial Networks
- Adversarial Generation of Natural Language
- On integral probability metrics, ϕ-divergences and binary classification
- On distinguishability criteria for estimating generative models
- Improving Generalization and Stability of Generative Adversarial Networks
- Geometric GAN
- Generative adversarial network-based glottal waveform model for statistical parametric speech synthesis
- A Probe Towards Understanding GAN and VAE Models
- The Six Fronts of the Generative Adversarial Networks