Use of symmetric kernels for convolutional neural networks

05/23/2018
by   Viacheslav Dudar, et al.
0

At this work we introduce horizontally symmetric convolutional kernels for CNNs which make the network output invariant to horizontal flips of the image. We also study other types of symmetric kernels which lead to vertical flip invariance, and approximate rotational invariance. We show that usage of such kernels acts as regularizer, and improves generalization of the convolutional neural networks at the cost of more complicated training process.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
11/24/2017

Efficient and Invariant Convolutional Neural Networks for Dense Prediction

Convolutional neural networks have shown great success on feature extrac...
research
06/12/2014

Convolutional Kernel Networks

An important goal in visual recognition is to devise image representatio...
research
03/20/2018

Dynamic Sampling Convolutional Neural Networks

We present Dynamic Sampling Convolutional Neural Networks (DSCNN), where...
research
01/08/2023

Equivariant and Steerable Neural Networks: A review with special emphasis on the symmetric group

Convolutional neural networks revolutionized computer vision and natrual...
research
03/03/2022

Symmetry Structured Convolutional Neural Networks

We consider Convolutional Neural Networks (CNNs) with 2D structured feat...
research
03/20/2019

Convolution with even-sized kernels and symmetric padding

Compact convolutional neural networks gain efficiency mainly through dep...
research
10/19/2020

ERIC: Extracting Relations Inferred from Convolutions

Our main contribution is to show that the behaviour of kernels across mu...

Please sign up or login with your details

Forgot password? Click here to reset