Omni-GAN: On the Secrets of cGANs and Beyond

26 Nov 2020 Peng Zhou Lingxi Xie Bingbing Ni Cong Geng Qi Tian

The conditional generative adversarial network (cGAN) is a powerful tool of generating high-quality images, but existing approaches mostly suffer unsatisfying performance or the risk of mode collapse. This paper presents Omni-GAN, a variant of cGAN that reveals the devil in designing a proper discriminator for training the model... (read more)

PDF Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK BENCHMARK
Conditional Image Generation ImageNet 128x128 Omni-GAN FID 8.30 # 6
Inception score 190.94 # 2
Conditional Image Generation ImageNet 128x128 Omni-INR-GAN FID 6.53 # 3
Inception score 262.85 # 1

Methods used in the Paper


METHOD TYPE
Dense Connections
Feedforward Networks
Non-Local Operation
Image Feature Extractors
Feedforward Network
Feedforward Networks
ReLU
Activation Functions
Residual Connection
Skip Connections
1x1 Convolution
Convolutions
Dot-Product Attention
Attention Mechanisms
Batch Normalization
Normalization
Early Stopping
Regularization
Non-Local Block
Image Model Blocks
Projection Discriminator
Discriminators
Residual Block
Skip Connection Blocks
Softmax
Output Functions
Linear Layer
Feedforward Networks
Spectral Normalization
Normalization
Adam
Stochastic Optimization
Conditional Batch Normalization
Normalization
GAN Hinge Loss
Loss Functions
Convolution
Convolutions
SAGAN Self-Attention Module
Attention Modules
TTUR
Optimization
SAGAN
Generative Adversarial Networks
Truncation Trick
Latent Variable Sampling
Off-Diagonal Orthogonal Regularization
Regularization
BigGAN
Generative Models
Weight Decay
Regularization