In this paper, we consider the problem of descriptors construction for the
task of content-based image retrieval using deep neural networks. The idea of
neural codes, based on fully connected layers activations, is extended by
incorporating the information contained in convolutional layers. It is known
that the total number of neurons in the convolutional part of the network is
large and the majority of them have little influence on the final
classification decision. Therefore, in the paper we propose a novel algorithm
that allows us to extract the most significant neuron activations and utilize
this information to construct effective descriptors. The descriptors consisting
of values taken from both the fully connected and convolutional layers
perfectly represent the whole image content. The images retrieved using these
descriptors match semantically very well to the query image, and also they are
similar in other secondary image characteristics, like background, textures or
color distribution. These features of the proposed descriptors are verified
experimentally based on the IMAGENET1M dataset using the VGG16 neural network.
Description
[2007.06624] A new approach to descriptors generation for image retrieval by analyzing activations of deep neural network layers
%0 Journal Article
%1 staszewski2020approach
%A Staszewski, Paweł
%A Jaworski, Maciej
%A Cao, Jinde
%A Rutkowski, Leszek
%D 2020
%J arXiv:2007.06624 cs
%K automatisches_indexieren bilder
%T A new approach to descriptors generation for image retrieval by
analyzing activations of deep neural network layers
%U http://arxiv.org/abs/2007.06624
%X In this paper, we consider the problem of descriptors construction for the
task of content-based image retrieval using deep neural networks. The idea of
neural codes, based on fully connected layers activations, is extended by
incorporating the information contained in convolutional layers. It is known
that the total number of neurons in the convolutional part of the network is
large and the majority of them have little influence on the final
classification decision. Therefore, in the paper we propose a novel algorithm
that allows us to extract the most significant neuron activations and utilize
this information to construct effective descriptors. The descriptors consisting
of values taken from both the fully connected and convolutional layers
perfectly represent the whole image content. The images retrieved using these
descriptors match semantically very well to the query image, and also they are
similar in other secondary image characteristics, like background, textures or
color distribution. These features of the proposed descriptors are verified
experimentally based on the IMAGENET1M dataset using the VGG16 neural network.
@article{staszewski2020approach,
abstract = {In this paper, we consider the problem of descriptors construction for the
task of content-based image retrieval using deep neural networks. The idea of
neural codes, based on fully connected layers activations, is extended by
incorporating the information contained in convolutional layers. It is known
that the total number of neurons in the convolutional part of the network is
large and the majority of them have little influence on the final
classification decision. Therefore, in the paper we propose a novel algorithm
that allows us to extract the most significant neuron activations and utilize
this information to construct effective descriptors. The descriptors consisting
of values taken from both the fully connected and convolutional layers
perfectly represent the whole image content. The images retrieved using these
descriptors match semantically very well to the query image, and also they are
similar in other secondary image characteristics, like background, textures or
color distribution. These features of the proposed descriptors are verified
experimentally based on the IMAGENET1M dataset using the VGG16 neural network.},
added-at = {2020-07-16T17:52:15.000+0200},
author = {Staszewski, Paweł and Jaworski, Maciej and Cao, Jinde and Rutkowski, Leszek},
biburl = {https://www.bibsonomy.org/bibtex/28f215e4ae5dfe119280ee244b9c1fb18/lepsky},
description = {[2007.06624] A new approach to descriptors generation for image retrieval by analyzing activations of deep neural network layers},
interhash = {3d8d79fb2bd23eadeb8a372c7ee85425},
intrahash = {8f215e4ae5dfe119280ee244b9c1fb18},
journal = {arXiv:2007.06624 [cs]},
keywords = {automatisches_indexieren bilder},
note = {cite arxiv:2007.06624Comment: 8},
timestamp = {2020-07-16T17:52:15.000+0200},
title = {A new approach to descriptors generation for image retrieval by
analyzing activations of deep neural network layers},
url = {http://arxiv.org/abs/2007.06624},
year = 2020
}