InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets

This paper describes InfoGAN, an information-theoretic extension to the Generative Adversarial Network that is able to learn disentangled representations in a completely unsupervised manner. InfoGAN is a generative adversarial network that also maximizes the mutual information between a small subset of the latent variables and the observation... (read more)

PDF Abstract NeurIPS 2016 PDF NeurIPS 2016 Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK BENCHMARK
Unsupervised MNIST MNIST InfoGAN Accuracy 95 # 7
Unsupervised Image Classification MNIST InfoGAN Accuracy 95 # 6

Results from Other Papers


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK SOURCE PAPER COMPARE
Image Generation CUB 128 x 128 InfoGAN FID 13.20 # 2
Inception score 47.32 # 2
Image Generation Stanford Cars InfoGAN FID 17.63 # 2
Inception score 28.62 # 2
Image Generation Stanford Dogs InfoGAN FID 29.34 # 2
Inception score 43.16 # 2

Methods used in the Paper


METHOD TYPE
Dense Connections
Feedforward Networks
Softmax
Output Functions
ReLU
Activation Functions
Feedforward Network
Feedforward Networks
Leaky ReLU
Activation Functions
Sigmoid Activation
Activation Functions
Tanh Activation
Activation Functions
Adam
Stochastic Optimization
Batch Normalization
Normalization
Convolution
Convolutions
GAN
Generative Models
InfoGAN
Generative Models