paper-with-me

Papers

Iterative Normalization: Beyond Standardization towards Efficient Whitening

2019-04-06 · CVPR 2019 6 · Lei Huang, Yi Zhou, Fan Zhu, Li Liu, Ling Shao

Batch Normalization (BN) is ubiquitously employed for accelerating neural network training and improving the generalization capability by performing standardization within mini-batches. Decorrelated Batch Normalization (DBN) further boosts the above effectiveness by whitening. However, DBN relies heavily on either a large batch size, or eigen-decomposition that suffers from poor efficiency on GPUs. We propose Iterative Normalization (IterNorm), which employs Newton's iterations for much more efficient whitening, while simultaneously avoiding the eigen-decomposition. Furthermore, we develop a comprehensive study to show IterNorm has better trade-off between optimization and generalization, with theoretical and experimental support. To this end, we exclusively introduce Stochastic Normalization Disturbance (SND), which measures the inherent stochastic uncertainty of samples when applied to normalization operations. With the support of SND, we provide natural explanations to several phenomena from the perspective of optimization, e.g., why group-wise whitening of DBN generally outperforms full-whitening and why the accuracy of BN degenerates with reduced batch sizes. We demonstrate the consistently improved performance of IterNorm with extensive experiments on CIFAR-10 and ImageNet over BN and DBN.

📄 PDF Abstract BibTeX arXiv:1904.03441

Code (5)

huangleiBuaa/IterNorm 공식 구현 pytorch
XingangPan/Switchable-Whitening pytorch
bhneo/DecorrelatedBN_tf tf
bhneo/decorrelated_bn tf
huangleiBuaa/IterNorm-pytorch pytorch

Tasks

Robust Object Detection

Methods 이 논문이 사용한 방법론

ZCA Whitening ZCA Whitening is an image preprocessing method that leads to a transformation of data such that the covariance matrix $\Sigma$ is the identity matrix, leading to decorrelated…
Decorrelated Batch Normalization Decorrelated Batch Normalization (DBN) is a normalization technique which not just centers and scales activations but whitens them. ZCA whitening instead of…
Batch Normalization 설명 없음

Similar Papers 제목 키워드 기반

Switchable Whitening for Deep Representation Learning

2019-04-22 · ICCV 2019 10 · Xingang Pan, Xiaohang Zhan, Jianping Shi, Xiaoou Tang 외

Normalization methods are essential components in convolutional neural networks (CNNs). They either standardize or whiten data using statistics estimated in predefined sets of pixels. Unlike existing works that design no…

Domain Adaptationimage-classificationImage ClassificationRepresentation Learning+3

Group Whitening: Balancing Learning Efficiency and Representational Capacity

2020-09-28 · CVPR 2021 1 · Lei Huang, Yi Zhou, Li Liu, Fan Zhu 외

Batch normalization (BN) is an important technique commonly incorporated into deep learning models to perform standardization within mini-batches. The merits of BN in improving a model's learning efficiency can be furthe…

Stochastic Whitening Batch Normalization

2021-06-03 · CVPR 2021 1 · Shengdong Zhang, Ehsan Nezhadarya, Homa Fashandi, Jiayi Liu 외

Batch Normalization (BN) is a popular technique for training Deep Neural Networks (DNNs). BN uses scaling and shifting to normalize activations of mini-batches to accelerate convergence and improve generalization. The re…

image-classificationImage Classification

An Investigation into the Stochasticity of Batch Whitening

2020-03-27 · CVPR 2020 6 · Lei Huang, Lei Zhao, Yi Zhou, Fan Zhu 외

Batch Normalization (BN) is extensively employed in various network architectures by performing standardization within mini-batches. A full understanding of the process has been a central target in the deep learning comm…

Attribute

Improving Generalization of Batch Whitening by Convolutional Unit Optimization

2021-08-24 · ICCV 2021 10 · Yooshin Cho, Hanbyel Cho, Youngsoo Kim, Junmo Kim

Batch Whitening is a technique that accelerates and stabilizes training by transforming input features to have a zero mean (Centering) and a unit variance (Scaling), and by removing linear correlation between channels (D…

image-classificationImage Classification