Yann LeCun described it as "the most interesting idea of machine learning in the past 10 years." Of course, this compliment from a distinguished researcher in the field of deep learning is always a good advertisement for the topic we are talking about!
Generating Confrontation Network (GAN) Beginner's Guide – with code
generateConfrontation networkIntroduction to GANs for beginners, principle explanation-with code
A picture of everything! NVIDIA pushes super image conversion artifact, small sample one second cat changes dog
For a while, NVIDIA's StyleGANIt can be said that it is a fire, and recently made a big move! In the past, image-to-image conversion required a large number of images for training samples, but in this work of Nvidia, image-to-image conversion can be done with only a small sample (the code is open source)!
Graduating in GAN: From understanding the generation of confrontation networks to running your own network
Read the generative confrontational network (GANResearch and evaluate how to develop and then implement your own GAN to generate handwritten numbers
"Detailed GAN" Why training is so difficult to fight against the network!
This article isGANAs part of the series, we will study why training is so elusive. Through this research, we learned some basic questions that drive the direction of many researchers. We will study some differences so that we know where the research might go. Before studying these issues, let's quickly review the GAN equation.
The rise of the generative confrontation network (Gan)
Before 5, GANs began a revolution in deep learning. This revolution has produced some major technological breakthroughs. This article will introduce in detail GAN History and the effects of different models.
As of 2019, artificial intelligence (AI) is an exciting 5 breakthrough
Compared to other fields, machine learning/artificial intelligence now seems to have a super-funny development with higher frequency.
MirrorGAN was born! Zhejiang University and other papers propose a new text-image framework to refresh the COCO record
Researchers at universities such as Zhejiang University and Sydney University proposed MirrorGAN, as a global-local attention and semantically maintained text-image-text framework, addresses semantic consistency between textual descriptions and visual content, and refreshes records on COCO datasets.
GANs Tens, the first article of safety
The generation of anti-neural networks (GANs) is the key to the further development of deep learning, and it has great application prospects in many fields. But the prosperity of GANs also needs to cross the two mountains of hardware and framework.
PyTorch-based GAN framework TorchGAN: Easily customize GAN projects with architecture-level APIs
TorchGAN It is based on PyTorch's GAN design development framework. The framework is designed to provide building blocks for popular GANs and allows for customization of cutting-edge research.