Detecting Overfitting of Deep Generative Networks via Latent Recovery
This addresses the issue of overfitting detection for researchers and practitioners using deep generative models, particularly in image generation, and is incremental as it builds on existing overfitting analysis methods.
The paper tackles the problem of detecting overfitting in deep generative networks by analyzing reconstruction errors of training versus validation images, finding that overfitting is not detectable in pure GAN models but is in hybrid adversarial loss models, and that standard GAN evaluation metrics fail to capture memorization.
State of the art deep generative networks are capable of producing images with such incredible realism that they can be suspected of memorizing training images. It is why it is not uncommon to include visualizations of training set nearest neighbors, to suggest generated images are not simply memorized. We demonstrate this is not sufficient and motivates the need to study memorization/overfitting of deep generators with more scrutiny. This paper addresses this question by i) showing how simple losses are highly effective at reconstructing images for deep generators ii) analyzing the statistics of reconstruction errors when reconstructing training and validation images, which is the standard way to analyze overfitting in machine learning. Using this methodology, this paper shows that overfitting is not detectable in the pure GAN models proposed in the literature, in contrast with those using hybrid adversarial losses, which are amongst the most widely applied generative methods. The paper also shows that standard GAN evaluation metrics fail to capture memorization for some deep generators. Finally, the paper also shows how off-the-shelf GAN generators can be successfully applied to face inpainting and face super-resolution using the proposed reconstruction method, without hybrid adversarial losses.