Please use this identifier to cite or link to this item: http://hdl.handle.net/1942/31322
Title: A novel feature representation: Aggregating convolution kernels for image retrieval
Authors: WANG, Qi 
Lai, Jinxing
CLAESEN, Luc 
Yang, Zhenguo
Lei, Liang
Liu, Wenyin
Issue Date: 2020
Publisher: PERGAMON-ELSEVIER SCIENCE LTD
Source: Neural networks (Print), 130 (October 2020) , p. 1 -10
Abstract: Activated hidden unites in convolutional neural networks (CNNs), known as feature maps, dominate image representation, which is compact and discriminative. For ultra-large data sets, high dimensional feature maps in float format not only result in high computational complexity, but also occupy massive memory space. To this end, a new image representation by aggregating convolution kernels (ACK) is proposed, where some convolution kernels capturing certain patterns are activated. The top-n index numbers of the convolution kernels are extracted directly as image representation in discrete integer values, which rebuild relationship between convolution kernels and image. Furthermore, a distance measurement is defined from the perspective of ordered sets to calculate position-sensitive similarities between image representations. Extensive experiments conducted on Oxford Buildings, Paris, and Holidays, etc., manifest that the proposed ACK achieves competitive performance on image retrieval with much lower computational cost, outperforming the ones using feature maps for image representation.
Keywords: Image Retrieval;Image Representation;Feature Aggregating;Distance Measurement;Convolutional Neural Networks
Document URI: http://hdl.handle.net/1942/31322
ISSN: 0893-6080
e-ISSN: 1879-2782
DOI: 10.1016/j.neunet.2020.06.010
ISI #: WOS:000567813200001
Rights: 2020 Elsevier Ltd.
Category: A1
Type: Journal Contribution
Validations: ecoom 2021
Appears in Collections:Research publications

Files in This Item:
File Description SizeFormat 
WangQ_2020.pdf
  Restricted Access
Published version2.37 MBAdobe PDFView/Open    Request a copy
A Novel Feature Representation Aggregating Convolution Kernels for Image Retrieval.pdfPeer-reviewed author version12.55 MBAdobe PDFView/Open
Show full item record

WEB OF SCIENCETM
Citations

16
checked on Mar 29, 2024

Page view(s)

68
checked on Sep 7, 2022

Download(s)

16
checked on Sep 7, 2022

Google ScholarTM

Check

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.