NIPS 2016 Tutorial: Generative Adversarial Networks

By Ian J. Goodfellow. 2016

1,821 citations recorded by the source corpus.

This report summarizes the tutorial presented by the author at NIPS 2016 on generative adversarial networks (GANs). The tutorial describes: (1) Why generative modeling is a topic worth studying, (2) how generative models work, and how GANs compare to other generative models, (3) the details of how GANs work, (4) research frontiers in GANs, and (5) state-of-the-art image models that combine GANs with other methods. Finally, the tutorial contains three exercises for readers to complete, and the solutions to these exercises.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Generative Models

How researchers see this paper

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

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Generative Models

Citations in the corpus

Jukebox: A Generative Model for Music (2020, background): “Generative modeling in deep learning: Generative models aim to learn the distribution of data by either explicitly by modeling the distribution or implicitly by constructing means to sample from it (Goodfellow, 2016).”
Imperfect ImaGANation: Implications of GANs Exacerbating Biases on Facial Data Augmentation and Snapchat Selfie Lenses (2020, background): “While mode collapse is a well-studied phenomenon (Grnarova et al., 2018; Goodfellow, 2016; Che et al., 2017; Arora et al., 2017) and several GAN variants have been developed to alleviate its effects (Metz et al., 2017; Srivastava et al., 2017; Arjovsky et al., 2017; Miyato et al., 2018; Tolstikhin…”
Transformers as Soft Reasoners over Language (2020, background): “Ad-versarial approaches to generation, where the generator learns to create theories that are hard for a partially trained model, may be useful in this context, e.g., [ Kalyan et al. , 2019; Goodfellow, 2016 ] .”
A Tight and Unified Analysis of Extragradient for a Whole Spectrum of Differentiable Games (2019, methodology): “The presence of rotational dynamics prevent standard algorithms such as the gradient method to converge on simple bilinear examples (Goodfellow, 2016; Balduzzi et al., 2018).”
Painless Stochastic Gradient: Interpolation, Line-Search, and Convergence Rates (2019, methodology): “Furthermore, the interpolation property can be used to improve the convergence for a bilinear saddle-point problem [25, 89, 56, 26].”
Many Paths to Equilibrium: GANs Do Not Need to Decrease a Divergence At Every Step (2017, background): “Goodfellow et al. [8] use this observation to establish that there is a unique Nash equilibrium in function space corresponding to p data = p model .”
Coulomb GANs: Provably Optimal Nash Equilibria via Potential Fields (2017, background): “However batch normalization introduces ﬂuctuations of normalizing constants which can be harmful [26, 15].”
Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks (2017, methodology): “Qualitative results are presented on several tasks where paired training data does not exist, including collection style transfer, object transfiguration, season transfer, photo enhancement, etc. Quantitative comparisons against several prior methods demonstrate the superiority of our approach.”
Unsupervised Machine Translation Using Monolingual Corpora Only (2017, methodology): “Following Goodfellow (2016), we include a smoothing coefﬁcient s = 0 .”
Word Translation Without Parallel Data (2017, methodology): “As suggested by Goodfellow (2016), we include a smoothing coefﬁcient s = 0 .”
BEGAN: Boundary Equilibrium Generative Adversarial Networks (2017, background): “Balancing the convergence of the discriminator and of the generator is a challenge: frequently the discriminator wins too easily at the beginning of training [6].”
GANs Trained by a Two Time-Scale Update Rule Converge to a Nash Equilibrium (2017, background): “Since training GANs is a game and its solution is a Nash equilibrium, gradient descent may fail to converge [48, 17, 19].”
A Variational Inequality Perspective on Generative Adversarial Networks (2018, methodology): “Yet it is known that for some games (Goodfellow, 2016, §8.2) SGD exhibits oscillatory behavior and fails to converge.”
Adversarial Examples: Attacks and Defenses for Deep Learning (2017, background): “We name this approach Natural GAN.”
A Survey on Data Collection for Machine Learning: A Big Data - AI Integration Perspective (2018, methodology): “Synthetic Data Generative Adversarial Networks [57], [58], [59], [60], [61], [62] Policies [63], [64] Image [65], [66], [67], [68], [69], [70], [71] Text [72], [73], [74]”