Visual Semantic Image Recommendation
Image recommendation is an essential component of the modern online image sharing applications (e.g., Flickr), aiming to provide users with interesting images for further exploration. However, most existing approaches tend to treat the image in question as a single object, ignoring the important sem...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2019-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/8648433/ |
id |
doaj-bc4d73d3f3b64cefa699abcca61f0531 |
---|---|
record_format |
Article |
spelling |
doaj-bc4d73d3f3b64cefa699abcca61f05312021-03-29T22:58:14ZengIEEEIEEE Access2169-35362019-01-017334243343310.1109/ACCESS.2019.29003968648433Visual Semantic Image RecommendationGuibing Guo0Yuan Meng1https://orcid.org/0000-0001-6420-5708Yongfeng Zhang2Chunyan Han3Yanjie Li4Department of Software, Northeastern University, Shenyang, ChinaDepartment of Software, Northeastern University, Shenyang, ChinaDepartment of Computer Science, Rutgers University, Piscataway, NJ, USADepartment of Software, Northeastern University, Shenyang, ChinaDepartment of Software, Northeastern University, Shenyang, ChinaImage recommendation is an essential component of the modern online image sharing applications (e.g., Flickr), aiming to provide users with interesting images for further exploration. However, most existing approaches tend to treat the image in question as a single object, ignoring the important semantics of the sub-objects within the image. The loss of these semantic objects may lead to the misunderstanding of the user preference toward an image. In this paper, we propose a novel pairwise preference model, called Visual Semantic Model (VSM), to address this issue for a better recommendation. Specifically, we model the image representation by combining the feature embeddings of the fine-grained image objects, the weights of which may be distinct for different users. Then, we enhance the user modeling by taking into account the interacted images along with their relative importance. Two attention networks on both object and image levels are adapted to compute the weights of objects and images, respectively. The experimental results on the Flickr dataset show that our VSM model achieves significant improvements (around 9.18% on average in terms of Precision@5) over the state-of-the-art approaches in terms of the recommendation accuracy.https://ieeexplore.ieee.org/document/8648433/Image recommendationsemantic objectsattention networks |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Guibing Guo Yuan Meng Yongfeng Zhang Chunyan Han Yanjie Li |
spellingShingle |
Guibing Guo Yuan Meng Yongfeng Zhang Chunyan Han Yanjie Li Visual Semantic Image Recommendation IEEE Access Image recommendation semantic objects attention networks |
author_facet |
Guibing Guo Yuan Meng Yongfeng Zhang Chunyan Han Yanjie Li |
author_sort |
Guibing Guo |
title |
Visual Semantic Image Recommendation |
title_short |
Visual Semantic Image Recommendation |
title_full |
Visual Semantic Image Recommendation |
title_fullStr |
Visual Semantic Image Recommendation |
title_full_unstemmed |
Visual Semantic Image Recommendation |
title_sort |
visual semantic image recommendation |
publisher |
IEEE |
series |
IEEE Access |
issn |
2169-3536 |
publishDate |
2019-01-01 |
description |
Image recommendation is an essential component of the modern online image sharing applications (e.g., Flickr), aiming to provide users with interesting images for further exploration. However, most existing approaches tend to treat the image in question as a single object, ignoring the important semantics of the sub-objects within the image. The loss of these semantic objects may lead to the misunderstanding of the user preference toward an image. In this paper, we propose a novel pairwise preference model, called Visual Semantic Model (VSM), to address this issue for a better recommendation. Specifically, we model the image representation by combining the feature embeddings of the fine-grained image objects, the weights of which may be distinct for different users. Then, we enhance the user modeling by taking into account the interacted images along with their relative importance. Two attention networks on both object and image levels are adapted to compute the weights of objects and images, respectively. The experimental results on the Flickr dataset show that our VSM model achieves significant improvements (around 9.18% on average in terms of Precision@5) over the state-of-the-art approaches in terms of the recommendation accuracy. |
topic |
Image recommendation semantic objects attention networks |
url |
https://ieeexplore.ieee.org/document/8648433/ |
work_keys_str_mv |
AT guibingguo visualsemanticimagerecommendation AT yuanmeng visualsemanticimagerecommendation AT yongfengzhang visualsemanticimagerecommendation AT chunyanhan visualsemanticimagerecommendation AT yanjieli visualsemanticimagerecommendation |
_version_ |
1724190472121876480 |