MS-RMAC: Multiscale Regional Maximum Activation of Convolutions for Image Retrieval

Recent works have demonstrated that image descriptors produced by convolutional feature maps provide state-of-the-art performance for image retrieval and classification problems. However, features from a single convolutional layer are not robust enough for shape deformation, scale variation, and hea...

Full description

Saved in:
Bibliographic Details
Published in:IEEE signal processing letters 2017-05, Vol.24 (5), p.609-613
Main Authors: Li, Yang, Xu, Yulong, Wang, Jiabao, Miao, Zhuang, Zhang, Yafei
Format: Article
Language:English
Subjects:
Convolution
Convolutional codes
Convolutional neural network (CNN)
Feature extraction
Image retrieval
multiscale (MS) feature
Neural networks
Robustness
Signal processing algorithms
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Recent works have demonstrated that image descriptors produced by convolutional feature maps provide state-of-the-art performance for image retrieval and classification problems. However, features from a single convolutional layer are not robust enough for shape deformation, scale variation, and heavy occlusion. In this letter, we present a simple and straightforward approach for extracting multiscale (MS) regional maximum activation of convolutions features from different layers of the convolutional neural network. And we also propose aggregating MS features into a single vector by a parameter-free hedge method for image retrieval. Extensive experimental results on three challenging benchmark datasets indicate that the proposed method achieved outstanding performance against state-of-the-art methods.
ISSN:1070-9908
1558-2361
DOI:10.1109/LSP.2017.2665522