Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
Alec Radford, Luke Metz, Soumith Chintala · arXiv (Cornell University) · 2015
In recent years, supervised learning with convolutional networks (CNNs) has seen huge adoption in computer vision applications. Comparatively, unsupervised learning with CNNs has received less attention. In this work we hope to help bridge the gap between the success of CNNs for supervised learning and unsupervised learning. We introduce a class of CNNs called deep convolutional generative adversarial networks (DCGANs), that have certain architectural constraints, and demonstrate that they are a strong candidate for unsupervised learning. Training on various image datasets, we show convincing evidence that our deep convolutional adversarial pair learns a hierarchy of representations from object parts to scenes in both the generator and discriminator. Additionally, we use the learned features for novel tasks - demonstrating their applicability as general image representations.
Cited by 6,396 works. doi:10.48550/arxiv.1511.06434 Open-access PDF
Authors
References
- ImageNet: A large-scale hierarchical image database
- Adam: A Method for Stochastic Optimization
- Visualizing and Understanding Convolutional Networks
- Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
- GAN(Generative Adversarial Nets)
- Convolutional deep belief networks for scalable unsupervised learning of hierarchical representations
- Learning and Transferring Mid-level Image Representations Using Convolutional Neural Networks
- Stacked Denoising Autoencoders: Learning Useful Representations in a Deep Network with a Local Denoising Criterion
- Texture synthesis by non-parametric sampling
- Random search for hyper-parameter optimization