Ambient Hidden Space of Generative Adversarial Networks
This work addresses a specific technical bottleneck in GAN training for researchers in generative modeling, though it appears incremental.
The paper tackles the limitation of ambient modules being only applicable to the output space of generative adversarial networks (GANs) by extending them to the hidden space, providing uniqueness conditions and training strategies. It reports practical results on benchmark datasets.
Generative adversarial models are powerful tools to model structure in complex distributions for a variety of tasks. Current techniques for learning generative models require an access to samples which have high quality, and advanced generative models are applied to generate samples from noisy training data through ambient modules. However, the modules are only practical for the output space of the generator, and their application in the hidden space is not well studied. In this paper, we extend the ambient module to the hidden space of the generator, and provide the uniqueness condition and the corresponding strategy for the ambient hidden generator in the adversarial training process. We report the practicality of the proposed method on the benchmark dataset.