Understanding Regularization to Visualize Convolutional Neural Networks
arXiv:1805.00071
Abstract
Variational methods for revealing visual concepts learned by convolutional neural networks have gained significant attention during the last years. Being based on noisy gradients obtained via back-propagation such methods require the application of regularization strategies. We present a mathematical framework unifying previously employed regularization methods. Within this framework, we propose a novel technique based on Sobolev gradients which can be implemented via convolutions and does not require specialized numerical treatment, such as total variation regularization. The experiments performed on feature inversion and activation maximization demonstrate the benefit of a unified approach to regularization, such as sharper reconstructions via the proposed Sobolev filters and a better control over reconstructed scales.
References in corpus (9)
- Axiomatic Attribution for Deep Networks
- Striving for Simplicity: The All Convolutional Net
- Visualizing Deep Convolutional Neural Networks Using Natural Pre-Images
- Learning how to explain neural networks: PatternNet and PatternAttribution
- Synthesizing the preferred inputs for neurons in neural networks via deep generator networks
- Regularization for Deep Learning: A Taxonomy
- The (Un)reliability of saliency methods
- Plug & Play Generative Networks: Conditional Iterative Generation of Images in Latent Space
- A Taxonomy and Library for Visualizing Learned Features in Convolutional Neural Networks