Content Similarity Assessment for Group Photo Retrieval

碩士 === 國立中興大學 === 資訊科學與工程學系 === 103 === In this paper, we aim to achieve the group photo retrieval by assessing the content similarity of subjects in the group photos. We consider the spatial arrangements and gender information of people to assess the visual similarity between two group photos. For...

Full description

Bibliographic Details
Main Authors: Ru-Yun Hsu, 許茹雲
Other Authors: Chun-Rong Huang
Format: Others
Language:en_US
Published: 2015
Online Access:http://ndltd.ncl.edu.tw/handle/ww84vp
id ndltd-TW-103NCHU5394044
record_format oai_dc
spelling ndltd-TW-103NCHU53940442019-05-15T22:25:04Z http://ndltd.ncl.edu.tw/handle/ww84vp Content Similarity Assessment for Group Photo Retrieval 基於內容物相似度評估之團體照檢索 Ru-Yun Hsu 許茹雲 碩士 國立中興大學 資訊科學與工程學系 103 In this paper, we aim to achieve the group photo retrieval by assessing the content similarity of subjects in the group photos. We consider the spatial arrangements and gender information of people to assess the visual similarity between two group photos. For each group photo, we represent detected faces by using the spatial face context. Then, a graph matching with face orientation and gender information is performed to assess the similarity between two group photos. As shown in the experimental results, our method can retrieve group photos with similar orientations and genders of subjects compared to state-of-the-art methods. Chun-Rong Huang 黃春融 2015 學位論文 ; thesis 39 en_US
collection NDLTD
language en_US
format Others
sources NDLTD
description 碩士 === 國立中興大學 === 資訊科學與工程學系 === 103 === In this paper, we aim to achieve the group photo retrieval by assessing the content similarity of subjects in the group photos. We consider the spatial arrangements and gender information of people to assess the visual similarity between two group photos. For each group photo, we represent detected faces by using the spatial face context. Then, a graph matching with face orientation and gender information is performed to assess the similarity between two group photos. As shown in the experimental results, our method can retrieve group photos with similar orientations and genders of subjects compared to state-of-the-art methods.
author2 Chun-Rong Huang
author_facet Chun-Rong Huang
Ru-Yun Hsu
許茹雲
author Ru-Yun Hsu
許茹雲
spellingShingle Ru-Yun Hsu
許茹雲
Content Similarity Assessment for Group Photo Retrieval
author_sort Ru-Yun Hsu
title Content Similarity Assessment for Group Photo Retrieval
title_short Content Similarity Assessment for Group Photo Retrieval
title_full Content Similarity Assessment for Group Photo Retrieval
title_fullStr Content Similarity Assessment for Group Photo Retrieval
title_full_unstemmed Content Similarity Assessment for Group Photo Retrieval
title_sort content similarity assessment for group photo retrieval
publishDate 2015
url http://ndltd.ncl.edu.tw/handle/ww84vp
work_keys_str_mv AT ruyunhsu contentsimilarityassessmentforgroupphotoretrieval
AT xǔrúyún contentsimilarityassessmentforgroupphotoretrieval
AT ruyunhsu jīyúnèiróngwùxiāngshìdùpínggūzhītuántǐzhàojiǎnsuǒ
AT xǔrúyún jīyúnèiróngwùxiāngshìdùpínggūzhītuántǐzhàojiǎnsuǒ
_version_ 1719129423960080384