Sparsity Aware Normalization for GANs

• URL: http://arxiv.org/abs/2103.02458v1
• Date: Wed, 3 Mar 2021 15:05:18 GMT
• Title: Sparsity Aware Normalization for GANs
• Authors: Idan Kligvasser, Tomer Michaeli
• Abstract summary: Generative adversarial networks (GANs) are known to benefit from regularization or normalization of their critic (discriminator) network during training. In this paper, we analyze the popular spectral normalization scheme, find a significant drawback and introduce sparsity aware normalization (SAN), a new alternative approach for stabilizing GAN training.
• Score: 32.76828505875087
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Generative adversarial networks (GANs) are known to benefit from regularization or normalization of their critic (discriminator) network during training. In this paper, we analyze the popular spectral normalization scheme, find a significant drawback and introduce sparsity aware normalization (SAN), a new alternative approach for stabilizing GAN training. As opposed to other normalization methods, our approach explicitly accounts for the sparse nature of the feature maps in convolutional networks with ReLU activations. We illustrate the effectiveness of our method through extensive experiments with a variety of network architectures. As we show, sparsity is particularly dominant in critics used for image-to-image translation settings. In these cases our approach improves upon existing methods, in less training epochs and with smaller capacity networks, while requiring practically no computational overhead.

Related papers

• NL-CS Net: Deep Learning with Non-Local Prior for Image Compressive Sensing [7.600617428107161]
  Deep learning has been applied to compressive sensing (CS) of images successfully in recent years. This paper proposes a novel CS method using non-local prior which combines the interpretability of the traditional optimization methods with the speed of network-based methods, called NL-CS Net.
  arXiv  Detail & Related papers  (2023-05-06T02:34:28Z)
• Low-light Image Enhancement by Retinex Based Algorithm Unrolling and Adjustment [50.13230641857892]
  We propose a new deep learning framework for the low-light image enhancement (LIE) problem. The proposed framework contains a decomposition network inspired by algorithm unrolling, and adjustment networks considering both global brightness and local brightness sensitivity. Experiments on a series of typical LIE datasets demonstrated the effectiveness of the proposed method, both quantitatively and visually, as compared with existing methods.
  arXiv  Detail & Related papers  (2022-02-12T03:59:38Z)
• Self-Ensembling GAN for Cross-Domain Semantic Segmentation [107.27377745720243]
  This paper proposes a self-ensembling generative adversarial network (SE-GAN) exploiting cross-domain data for semantic segmentation. In SE-GAN, a teacher network and a student network constitute a self-ensembling model for generating semantic segmentation maps, which together with a discriminator, forms a GAN. Despite its simplicity, we find SE-GAN can significantly boost the performance of adversarial training and enhance the stability of the model.
  arXiv  Detail & Related papers  (2021-12-15T09:50:25Z)
• Convergence Analysis and Implicit Regularization of Feedback Alignment for Deep Linear Networks [27.614609336582568]
  We theoretically analyze the Feedback Alignment (FA) algorithm, an efficient alternative to backpropagation for training neural networks. We provide convergence guarantees with rates for deep linear networks for both continuous and discrete dynamics.
  arXiv  Detail & Related papers  (2021-10-20T22:57:03Z)
• Distribution Mismatch Correction for Improved Robustness in Deep Neural Networks [86.42889611784855]
  normalization methods increase the vulnerability with respect to noise and input corruptions. We propose an unsupervised non-parametric distribution correction method that adapts the activation distribution of each layer. In our experiments, we empirically show that the proposed method effectively reduces the impact of intense image corruptions.
  arXiv  Detail & Related papers  (2021-10-05T11:36:25Z)
• Embracing the Dark Knowledge: Domain Generalization Using Regularized Knowledge Distillation [65.79387438988554]
  Lack of generalization capability in the absence of sufficient and representative data is one of the challenges that hinder their practical application. We propose a simple, effective, and plug-and-play training strategy named Knowledge Distillation for Domain Generalization (KDDG) We find that both the richer dark knowledge" from the teacher network, as well as the gradient filter we proposed, can reduce the difficulty of learning the mapping.
  arXiv  Detail & Related papers  (2021-07-06T14:08:54Z)
• On the Explicit Role of Initialization on the Convergence and Implicit Bias of Overparametrized Linear Networks [1.0323063834827415]
  We present a novel analysis of single-hidden-layer linear networks trained under gradient flow. We show that the squared loss converges exponentially to its optimum. We derive a novel non-asymptotic upper-bound on the distance between the trained network and the min-norm solution.
  arXiv  Detail & Related papers  (2021-05-13T15:13:51Z)
• Local Critic Training for Model-Parallel Learning of Deep Neural Networks [94.69202357137452]
  We propose a novel model-parallel learning method, called local critic training. We show that the proposed approach successfully decouples the update process of the layer groups for both convolutional neural networks (CNNs) and recurrent neural networks (RNNs) We also show that trained networks by the proposed method can be used for structural optimization.
  arXiv  Detail & Related papers  (2021-02-03T09:30:45Z)
• Keep the Gradients Flowing: Using Gradient Flow to Study Sparse Network Optimization [16.85167651136133]
  We take a broader view of training sparse networks and consider the role of regularization, optimization and architecture choices on sparse models. We show that gradient flow in sparse networks can be improved by reconsidering aspects of the architecture design and the training regime.
  arXiv  Detail & Related papers  (2021-02-02T18:40:26Z)
• Normalization Techniques in Training DNNs: Methodology, Analysis and Application [111.82265258916397]
  Normalization techniques are essential for accelerating the training and improving the generalization of deep neural networks (DNNs) This paper reviews and comments on the past, present and future of normalization methods in the context of training.
  arXiv  Detail & Related papers  (2020-09-27T13:06:52Z)
• Fiedler Regularization: Learning Neural Networks with Graph Sparsity [6.09170287691728]
  We introduce a novel regularization approach for deep learning that incorporates and respects the underlying graphical structure of the neural network. We propose to use the Fiedler value of the neural network's underlying graph as a tool for regularization.
  arXiv  Detail & Related papers  (2020-03-02T16:19:33Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.