An official website of the United States government
Batch normalization (BN) is a popular and ubiquitous method in deep learning that has been shown to decrease training time and improve generalization performance of neural networks. Despite its success, BN is not theoretically well understood. It is not suitable for use with very small mini-batch sizes or online learning. In this paper, we propose a new method called Batch Normalization Preconditioning (BNP). Instead of applying normalization explicitly through a batch normalization layer as is done in BN, BNP applies normalization by conditioning the parameter gradients directly during training. This is designed to improve the Hessian matrix of the loss function and hence convergence during training. One benefit is that BNP is not constrained on the mini-batch size and works in the online learning setting. Furthermore, its connection to BN provides theoretical insights on how BN improves training and how BN is applied to special architectures such as convolutional neural networks. For a theoretical foundation, we also present a novel Hessian condition number based convergence theory for a locally convex but not strong-convex loss, which is applicable to networks with a scale-invariant property.
more »
« less
Training Faster by Separating Modes of Variation in Batch-normalized Models
Batch Normalization (BN) is essential to effectively train state-of-the-art deep Convolutional Neural Networks (CNN). It normalizes the layer outputs during training using the statistics of each mini-batch. BN accelerates training procedure by allowing to safely utilize large learning rates and alleviates the need for careful initialization of the parameters. In this work, we study BN from the viewpoint of Fisher kernels that arise from generative probability models. We show that assuming samples within a mini-batch are from the same probability density function, then BN is identical to the Fisher vector of a Gaussian distribution. That means batch normalizing transform can be explained in terms of kernels that naturally emerge from the probability density function that models the generative process of the underlying data distribution. Consequently, it promises higher discrimination power for the batch-normalized mini-batch. However, given the rectifying non-linearities employed in CNN architectures, distribution of the layer outputs show an asymmetric characteristic. Therefore, in order for BN to fully benefit from the aforementioned properties, we propose approximating underlying data distribution not with one, but a mixture of Gaussian densities. Deriving Fisher vector for a Gaussian Mixture Model (GMM), reveals that batch normalization can be improved by independently normalizing with respect to the statistics of disentangled sub-populations. We refer to our proposed soft piecewise version of batch normalization as Mixture Normalization (MN). Through extensive set of experiments on CIFAR-10 and CIFAR-100, using both a 5-layers deep CNN and modern Inception-V3 architecture, we show that mixture normalization reduces required number of gradient updates to reach the maximum test accuracy of the batch normalized model by ∼31%-47% across a variety of training scenarios. Replacing even a few BN modules with MN in the 48-layers deep Inception-V3 architecture is sufficient to not only obtain considerable training acceleration but also better final test accuracy. We show that similar observations are valid for 40 and 100-layers deep DenseNet architectures as well. We complement our study by evaluating the application of mixture normalization to the Generative Adversarial Networks (GANs), where "mode collapse" hinders the training process. We solely replace a few batch normalization layers in the generator with our proposed mixture normalization. Our experiments using Deep Convolutional GAN (DCGAN) on CIFAR-10 show that mixture normalized DCGAN not only provides an acceleration of ∼58% but also reaches lower (better) "Fréchet Inception Distance" (FID) of 33.35 compared to 37.56 of its batch normalized counterpart.
more »
« less
- Award ID(s):
- 1741431
- PAR ID:
- 10111646
- Date Published:
- Journal Name:
- IEEE Transactions on Pattern Analysis and Machine Intelligence
- ISSN:
- 0162-8828
- Page Range / eLocation ID:
- 1 to 1
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
We develop a measure for evaluating the performance of generative networks given two sets of images. A popular performance measure currently used to do this is the Fréchet Inception Distance (FID). FID assumes that images featurized using the penultimate layer of Inception-v3 follow a Gaussian distribution, an assumption which cannot be violated if we wish to use FID as a metric. However, we show that Inception-v3 features of the ImageNet dataset are not Gaussian; in particular, every single marginal is not Gaussian. To remedy this problem, we model the featurized images using Gaussian mixture models (GMMs) and compute the 2-Wasserstein distance restricted to GMMs. We define a performance measure, which we call WaM, on two sets of images by using Inception-v3 (or another classifier) to featurize the images, estimate two GMMs, and use the restricted 2-Wasserstein distance to compare the GMMs. We experimentally show the advantages of WaM over FID, including how FID is more sensitive than WaM to imperceptible image perturbations. By modelling the non-Gaussian features obtained from Inception-v3 as GMMs and using a GMM metric, we can more accurately evaluate generative network performance.more » « less
-
Model confidence or uncertainty is critical in autonomous systems as they directly tie to the safety and trustworthiness of the system. The quantification of uncertainty in the output decisions of deep neural networks (DNNs) is a challenging problem. The Bayesian framework enables the estimation of the predictive uncertainty by introducing probability distributions over the (unknown) network weights; however, the propagation of these high-dimensional distributions through multiple layers and non-linear transformations is mathematically intractable. In this work, we propose an extended variational inference (eVI) framework for convolutional neural network (CNN) based on tensor Normal distributions (TNDs) defined over convolutional kernels. Our proposed eVI framework propagates the first two moments (mean and covariance) of these TNDs through all layers of the CNN. We employ first-order Taylor series linearization to approximate the mean and covariances passing through the non-linear activations. The uncertainty in the output decision is given by the propagated covariance of the predictive distribution. Furthermore, we show, through extensive simulations on the MNIST and CIFAR-10 datasets, that the CNN becomes more robust to Gaussian noise and adversarial attacks.more » « less
-
Self-attention is key to the remarkable success of transformers in sequence modeling tasks including many applications in natural language processing and computer vision. Like neural network layers, these attention mechanisms are often developed by heuristics and experience. To provide a principled framework for constructing attention layers in transformers, we show that the self-attention corresponds to the support vector expansion derived from a support vector regression problem, whose primal formulation has the form of a neural network layer. Using our framework, we derive popular attention layers used in practice and propose two new attentions: 1) the Batch Normalized Attention (Attention-BN) derived from the batch normalization layer and 2) the Attention with Scaled Head (Attention-SH) derived from using less training data to fit the SVR model. We empirically demonstrate the advantages of the Attention-BN and Attention-SH in reducing head redundancy, increasing the model's accuracy, and improving the model's efficiency in a variety of practical applications including image and time-series classification.more » « less
-
Chen, Xinyun; Xie, Cihang; Shafahi, Ali; Li, Bo; Zhao, Ding; Goldstein, Tom; Song, Dawn (Ed.)The adversarial training paradigm has become the standard in training deep neural networks for robustness. Yet, it remains unstable, with the mechanisms driving this instability poorly understood. In this study, we discover that this instability is primarily driven by a non-smooth optimization landscape and an internal covariate shift phenomenon, and show that Batch Normalization (BN) can effectively mitigate both these issues. Further, we demonstrate that BN universally improves clean and robust performance across various defenses, datasets, and model types, with greater improvement on more difficult tasks. Finally, we confirm BN’s heterogeneous distribution issue with mixed-batch training and propose a solution.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript
- Journal Article:
- https://doi.org/10.1109/TPAMI.2019.2895781
