Normalized Convolutional Neural Network

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kim, Dongsuk, Lee, Geonhee, Lee, Myungjae, Kang, Shin Uk, Kim, Dongmin
Format: Preprint
Published: 2020
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912304136716288
author Kim, Dongsuk
Lee, Geonhee
Lee, Myungjae
Kang, Shin Uk
Kim, Dongmin
author_facet Kim, Dongsuk
Lee, Geonhee
Lee, Myungjae
Kang, Shin Uk
Kim, Dongmin
contents We introduce a Normalized Convolutional Neural Layer, a novel approach to normalization in convolutional networks. Unlike conventional methods, this layer normalizes the rows of the im2col matrix during convolution, making it inherently adaptive to sliced inputs and better aligned with kernel structures. This distinctive approach differentiates it from standard normalization techniques and prevents direct integration into existing deep learning frameworks optimized for traditional convolution operations. Our method has a universal property, making it applicable to any deep learning task involving convolutional layers. By inherently normalizing within the convolution process, it serves as a convolutional adaptation of Self-Normalizing Networks, maintaining their core principles without requiring additional normalization layers. Notably, in micro-batch training scenarios, it consistently outperforms other batch-independent normalization methods. This performance boost arises from standardizing the rows of the im2col matrix, which theoretically leads to a smoother loss gradient and improved training stability.
format Preprint
id arxiv_https___arxiv_org_abs_2005_05274
institution arXiv
publishDate 2020
record_format arxiv
spellingShingle Normalized Convolutional Neural Network
Kim, Dongsuk
Lee, Geonhee
Lee, Myungjae
Kang, Shin Uk
Kim, Dongmin
Computer Vision and Pattern Recognition
We introduce a Normalized Convolutional Neural Layer, a novel approach to normalization in convolutional networks. Unlike conventional methods, this layer normalizes the rows of the im2col matrix during convolution, making it inherently adaptive to sliced inputs and better aligned with kernel structures. This distinctive approach differentiates it from standard normalization techniques and prevents direct integration into existing deep learning frameworks optimized for traditional convolution operations. Our method has a universal property, making it applicable to any deep learning task involving convolutional layers. By inherently normalizing within the convolution process, it serves as a convolutional adaptation of Self-Normalizing Networks, maintaining their core principles without requiring additional normalization layers. Notably, in micro-batch training scenarios, it consistently outperforms other batch-independent normalization methods. This performance boost arises from standardizing the rows of the im2col matrix, which theoretically leads to a smoother loss gradient and improved training stability.
title Normalized Convolutional Neural Network
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2005.05274