Query Specific Rank Fusion for Image Retrieval
Recently two lines of image retrieval algorithms demonstrate excellent scalability: 1) local features indexed by a vocabulary tree, and 2) holistic features indexed by compact hashing codes. Although both of them are able to search visually similar images effectively, their retrieval precision may v...
Ausführliche Beschreibung
Gespeichert in:
| Autor*in: |
Shaoting Zhang [verfasserIn] |
|---|
| Format: |
Artikel |
|---|---|
| Sprache: |
Englisch |
| Erschienen: |
2015 |
|---|
| Schlagwörter: |
graph-based query specific fusion approach top candidate nearest neighborhoods |
|---|
| Systematik: |
|
|---|
| Übergeordnetes Werk: |
Enthalten in: IEEE transactions on pattern analysis and machine intelligence - Washington, DC : IEEE Computer Soc., 1979, 37(2015), 4, Seite 803-815 |
|---|---|
| Übergeordnetes Werk: |
volume:37 ; year:2015 ; number:4 ; pages:803-815 |
| Links: |
|---|
| DOI / URN: |
10.1109/TPAMI.2014.2346201 |
|---|
| Katalog-ID: |
OLC1968068260 |
|---|
| LEADER | 01000caa a2200265 4500 | ||
|---|---|---|---|
| 001 | OLC1968068260 | ||
| 003 | DE-627 | ||
| 005 | 20220220222517.0 | ||
| 007 | tu | ||
| 008 | 160206s2015 xx ||||| 00| ||eng c | ||
| 024 | 7 | |a 10.1109/TPAMI.2014.2346201 |2 doi | |
| 028 | 5 | 2 | |a PQ20160617 |
| 035 | |a (DE-627)OLC1968068260 | ||
| 035 | |a (DE-599)GBVOLC1968068260 | ||
| 035 | |a (PRQ)c2358-e0de399188db465f4c60726f613e212a686433c749aca598af354b16c83666070 | ||
| 035 | |a (KEY)0091660920150000037000400803queryspecificrankfusionforimageretrieval | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rakwb | ||
| 041 | |a eng | ||
| 082 | 0 | 4 | |a 620 |a 004 |q DNB |
| 084 | |a SQ 1100: |q AVZ |2 rvk | ||
| 084 | |a 54.74 |2 bkl | ||
| 100 | 0 | |a Shaoting Zhang |e verfasserin |4 aut | |
| 245 | 1 | 0 | |a Query Specific Rank Fusion for Image Retrieval |
| 264 | 1 | |c 2015 | |
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a ohne Hilfsmittel zu benutzen |b n |2 rdamedia | ||
| 338 | |a Band |b nc |2 rdacarrier | ||
| 520 | |a Recently two lines of image retrieval algorithms demonstrate excellent scalability: 1) local features indexed by a vocabulary tree, and 2) holistic features indexed by compact hashing codes. Although both of them are able to search visually similar images effectively, their retrieval precision may vary dramatically among queries. Therefore, combining these two types of methods is expected to further enhance the retrieval precision. However, the feature characteristics and the algorithmic procedures of these methods are dramatically different, which is very challenging for the feature-level fusion. This motivates us to investigate how to fuse the ordered retrieval sets, i.e., the ranks of images, given by multiple retrieval methods, to boost the retrieval precision without sacrificing their scalability. In this paper, we model retrieval ranks as graphs of candidate images and propose a graph-based query specific fusion approach, where multiple graphs are merged and reranked by conducting a link analysis on a fused graph. The retrieval quality of an individual method is measured on-the-fly by assessing the consistency of the top candidates' nearest neighborhoods. Hence, it is capable of adaptively integrating the strengths of the retrieval methods using local or holistic features for different query images. This proposed method does not need any supervision, has few parameters, and is easy to implement. Extensive and thorough experiments have been conducted on four public datasets, i.e., the UKbench, Corel-5K, Holidays and the large-scale San Francisco Landmarks datasets. Our proposed method has achieved very competitive performance, including state-of-the-art results on several data sets, e.g., the N-S score 3.83 for UKbench. | ||
| 650 | 4 | |a compact hashing codes | |
| 650 | 4 | |a UKbench public datasets | |
| 650 | 4 | |a image retrieval algorithms | |
| 650 | 4 | |a query specific rank fusion | |
| 650 | 4 | |a link analysis | |
| 650 | 4 | |a graph-based query specific fusion approach | |
| 650 | 4 | |a Scalability | |
| 650 | 4 | |a Holidays public datasets | |
| 650 | 4 | |a image fusion | |
| 650 | 4 | |a feature-level fusion | |
| 650 | 4 | |a Visualization | |
| 650 | 4 | |a Vocabulary | |
| 650 | 4 | |a multiple retrieval methods | |
| 650 | 4 | |a top candidate nearest neighborhoods | |
| 650 | 4 | |a ordered retrieval set fusion | |
| 650 | 4 | |a Fuses | |
| 650 | 4 | |a trees (mathematics) | |
| 650 | 4 | |a Image edge detection | |
| 650 | 4 | |a vocabulary tree | |
| 650 | 4 | |a image retrieval | |
| 650 | 4 | |a Corel-5K public datasets | |
| 650 | 4 | |a large-scale San Francisco landmarks datasets | |
| 650 | 4 | |a Image retrieval | |
| 650 | 4 | |a Methods | |
| 700 | 0 | |a Ming Yang |4 oth | |
| 700 | 1 | |a Cour, Timothee |4 oth | |
| 700 | 0 | |a Kai Yu |4 oth | |
| 700 | 1 | |a Metaxas, Dimitris N |4 oth | |
| 773 | 0 | 8 | |i Enthalten in |t IEEE transactions on pattern analysis and machine intelligence |d Washington, DC : IEEE Computer Soc., 1979 |g 37(2015), 4, Seite 803-815 |w (DE-627)129616397 |w (DE-600)244051-9 |w (DE-576)015114341 |x 0162-8828 |7 nnns |
| 773 | 1 | 8 | |g volume:37 |g year:2015 |g number:4 |g pages:803-815 |
| 856 | 4 | 1 | |u http://dx.doi.org/10.1109/TPAMI.2014.2346201 |3 Volltext |
| 856 | 4 | 2 | |u http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6873347 |
| 856 | 4 | 2 | |u http://www.ncbi.nlm.nih.gov/pubmed/26353295 |
| 856 | 4 | 2 | |u http://search.proquest.com/docview/1663517075 |
| 912 | |a GBV_USEFLAG_A | ||
| 912 | |a SYSFLAG_A | ||
| 912 | |a GBV_OLC | ||
| 912 | |a SSG-OLC-TEC | ||
| 912 | |a SSG-OLC-MAT | ||
| 912 | |a GBV_ILN_30 | ||
| 912 | |a GBV_ILN_70 | ||
| 912 | |a GBV_ILN_2014 | ||
| 912 | |a GBV_ILN_2410 | ||
| 936 | r | v | |a SQ 1100: |
| 936 | b | k | |a 54.74 |q AVZ |
| 951 | |a AR | ||
| 952 | |d 37 |j 2015 |e 4 |h 803-815 | ||
| author_variant |
s z sz |
|---|---|
| matchkey_str |
article:01628828:2015----::urseiirnfsofrmg |
| hierarchy_sort_str |
2015 |
| bklnumber |
54.74 |
| publishDate |
2015 |
| allfields |
10.1109/TPAMI.2014.2346201 doi PQ20160617 (DE-627)OLC1968068260 (DE-599)GBVOLC1968068260 (PRQ)c2358-e0de399188db465f4c60726f613e212a686433c749aca598af354b16c83666070 (KEY)0091660920150000037000400803queryspecificrankfusionforimageretrieval DE-627 ger DE-627 rakwb eng 620 004 DNB SQ 1100: AVZ rvk 54.74 bkl Shaoting Zhang verfasserin aut Query Specific Rank Fusion for Image Retrieval 2015 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier Recently two lines of image retrieval algorithms demonstrate excellent scalability: 1) local features indexed by a vocabulary tree, and 2) holistic features indexed by compact hashing codes. Although both of them are able to search visually similar images effectively, their retrieval precision may vary dramatically among queries. Therefore, combining these two types of methods is expected to further enhance the retrieval precision. However, the feature characteristics and the algorithmic procedures of these methods are dramatically different, which is very challenging for the feature-level fusion. This motivates us to investigate how to fuse the ordered retrieval sets, i.e., the ranks of images, given by multiple retrieval methods, to boost the retrieval precision without sacrificing their scalability. In this paper, we model retrieval ranks as graphs of candidate images and propose a graph-based query specific fusion approach, where multiple graphs are merged and reranked by conducting a link analysis on a fused graph. The retrieval quality of an individual method is measured on-the-fly by assessing the consistency of the top candidates' nearest neighborhoods. Hence, it is capable of adaptively integrating the strengths of the retrieval methods using local or holistic features for different query images. This proposed method does not need any supervision, has few parameters, and is easy to implement. Extensive and thorough experiments have been conducted on four public datasets, i.e., the UKbench, Corel-5K, Holidays and the large-scale San Francisco Landmarks datasets. Our proposed method has achieved very competitive performance, including state-of-the-art results on several data sets, e.g., the N-S score 3.83 for UKbench. compact hashing codes UKbench public datasets image retrieval algorithms query specific rank fusion link analysis graph-based query specific fusion approach Scalability Holidays public datasets image fusion feature-level fusion Visualization Vocabulary multiple retrieval methods top candidate nearest neighborhoods ordered retrieval set fusion Fuses trees (mathematics) Image edge detection vocabulary tree image retrieval Corel-5K public datasets large-scale San Francisco landmarks datasets Image retrieval Methods Ming Yang oth Cour, Timothee oth Kai Yu oth Metaxas, Dimitris N oth Enthalten in IEEE transactions on pattern analysis and machine intelligence Washington, DC : IEEE Computer Soc., 1979 37(2015), 4, Seite 803-815 (DE-627)129616397 (DE-600)244051-9 (DE-576)015114341 0162-8828 nnns volume:37 year:2015 number:4 pages:803-815 http://dx.doi.org/10.1109/TPAMI.2014.2346201 Volltext http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6873347 http://www.ncbi.nlm.nih.gov/pubmed/26353295 http://search.proquest.com/docview/1663517075 GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-MAT GBV_ILN_30 GBV_ILN_70 GBV_ILN_2014 GBV_ILN_2410 SQ 1100: 54.74 AVZ AR 37 2015 4 803-815 |
| spelling |
10.1109/TPAMI.2014.2346201 doi PQ20160617 (DE-627)OLC1968068260 (DE-599)GBVOLC1968068260 (PRQ)c2358-e0de399188db465f4c60726f613e212a686433c749aca598af354b16c83666070 (KEY)0091660920150000037000400803queryspecificrankfusionforimageretrieval DE-627 ger DE-627 rakwb eng 620 004 DNB SQ 1100: AVZ rvk 54.74 bkl Shaoting Zhang verfasserin aut Query Specific Rank Fusion for Image Retrieval 2015 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier Recently two lines of image retrieval algorithms demonstrate excellent scalability: 1) local features indexed by a vocabulary tree, and 2) holistic features indexed by compact hashing codes. Although both of them are able to search visually similar images effectively, their retrieval precision may vary dramatically among queries. Therefore, combining these two types of methods is expected to further enhance the retrieval precision. However, the feature characteristics and the algorithmic procedures of these methods are dramatically different, which is very challenging for the feature-level fusion. This motivates us to investigate how to fuse the ordered retrieval sets, i.e., the ranks of images, given by multiple retrieval methods, to boost the retrieval precision without sacrificing their scalability. In this paper, we model retrieval ranks as graphs of candidate images and propose a graph-based query specific fusion approach, where multiple graphs are merged and reranked by conducting a link analysis on a fused graph. The retrieval quality of an individual method is measured on-the-fly by assessing the consistency of the top candidates' nearest neighborhoods. Hence, it is capable of adaptively integrating the strengths of the retrieval methods using local or holistic features for different query images. This proposed method does not need any supervision, has few parameters, and is easy to implement. Extensive and thorough experiments have been conducted on four public datasets, i.e., the UKbench, Corel-5K, Holidays and the large-scale San Francisco Landmarks datasets. Our proposed method has achieved very competitive performance, including state-of-the-art results on several data sets, e.g., the N-S score 3.83 for UKbench. compact hashing codes UKbench public datasets image retrieval algorithms query specific rank fusion link analysis graph-based query specific fusion approach Scalability Holidays public datasets image fusion feature-level fusion Visualization Vocabulary multiple retrieval methods top candidate nearest neighborhoods ordered retrieval set fusion Fuses trees (mathematics) Image edge detection vocabulary tree image retrieval Corel-5K public datasets large-scale San Francisco landmarks datasets Image retrieval Methods Ming Yang oth Cour, Timothee oth Kai Yu oth Metaxas, Dimitris N oth Enthalten in IEEE transactions on pattern analysis and machine intelligence Washington, DC : IEEE Computer Soc., 1979 37(2015), 4, Seite 803-815 (DE-627)129616397 (DE-600)244051-9 (DE-576)015114341 0162-8828 nnns volume:37 year:2015 number:4 pages:803-815 http://dx.doi.org/10.1109/TPAMI.2014.2346201 Volltext http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6873347 http://www.ncbi.nlm.nih.gov/pubmed/26353295 http://search.proquest.com/docview/1663517075 GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-MAT GBV_ILN_30 GBV_ILN_70 GBV_ILN_2014 GBV_ILN_2410 SQ 1100: 54.74 AVZ AR 37 2015 4 803-815 |
| allfields_unstemmed |
10.1109/TPAMI.2014.2346201 doi PQ20160617 (DE-627)OLC1968068260 (DE-599)GBVOLC1968068260 (PRQ)c2358-e0de399188db465f4c60726f613e212a686433c749aca598af354b16c83666070 (KEY)0091660920150000037000400803queryspecificrankfusionforimageretrieval DE-627 ger DE-627 rakwb eng 620 004 DNB SQ 1100: AVZ rvk 54.74 bkl Shaoting Zhang verfasserin aut Query Specific Rank Fusion for Image Retrieval 2015 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier Recently two lines of image retrieval algorithms demonstrate excellent scalability: 1) local features indexed by a vocabulary tree, and 2) holistic features indexed by compact hashing codes. Although both of them are able to search visually similar images effectively, their retrieval precision may vary dramatically among queries. Therefore, combining these two types of methods is expected to further enhance the retrieval precision. However, the feature characteristics and the algorithmic procedures of these methods are dramatically different, which is very challenging for the feature-level fusion. This motivates us to investigate how to fuse the ordered retrieval sets, i.e., the ranks of images, given by multiple retrieval methods, to boost the retrieval precision without sacrificing their scalability. In this paper, we model retrieval ranks as graphs of candidate images and propose a graph-based query specific fusion approach, where multiple graphs are merged and reranked by conducting a link analysis on a fused graph. The retrieval quality of an individual method is measured on-the-fly by assessing the consistency of the top candidates' nearest neighborhoods. Hence, it is capable of adaptively integrating the strengths of the retrieval methods using local or holistic features for different query images. This proposed method does not need any supervision, has few parameters, and is easy to implement. Extensive and thorough experiments have been conducted on four public datasets, i.e., the UKbench, Corel-5K, Holidays and the large-scale San Francisco Landmarks datasets. Our proposed method has achieved very competitive performance, including state-of-the-art results on several data sets, e.g., the N-S score 3.83 for UKbench. compact hashing codes UKbench public datasets image retrieval algorithms query specific rank fusion link analysis graph-based query specific fusion approach Scalability Holidays public datasets image fusion feature-level fusion Visualization Vocabulary multiple retrieval methods top candidate nearest neighborhoods ordered retrieval set fusion Fuses trees (mathematics) Image edge detection vocabulary tree image retrieval Corel-5K public datasets large-scale San Francisco landmarks datasets Image retrieval Methods Ming Yang oth Cour, Timothee oth Kai Yu oth Metaxas, Dimitris N oth Enthalten in IEEE transactions on pattern analysis and machine intelligence Washington, DC : IEEE Computer Soc., 1979 37(2015), 4, Seite 803-815 (DE-627)129616397 (DE-600)244051-9 (DE-576)015114341 0162-8828 nnns volume:37 year:2015 number:4 pages:803-815 http://dx.doi.org/10.1109/TPAMI.2014.2346201 Volltext http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6873347 http://www.ncbi.nlm.nih.gov/pubmed/26353295 http://search.proquest.com/docview/1663517075 GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-MAT GBV_ILN_30 GBV_ILN_70 GBV_ILN_2014 GBV_ILN_2410 SQ 1100: 54.74 AVZ AR 37 2015 4 803-815 |
| allfieldsGer |
10.1109/TPAMI.2014.2346201 doi PQ20160617 (DE-627)OLC1968068260 (DE-599)GBVOLC1968068260 (PRQ)c2358-e0de399188db465f4c60726f613e212a686433c749aca598af354b16c83666070 (KEY)0091660920150000037000400803queryspecificrankfusionforimageretrieval DE-627 ger DE-627 rakwb eng 620 004 DNB SQ 1100: AVZ rvk 54.74 bkl Shaoting Zhang verfasserin aut Query Specific Rank Fusion for Image Retrieval 2015 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier Recently two lines of image retrieval algorithms demonstrate excellent scalability: 1) local features indexed by a vocabulary tree, and 2) holistic features indexed by compact hashing codes. Although both of them are able to search visually similar images effectively, their retrieval precision may vary dramatically among queries. Therefore, combining these two types of methods is expected to further enhance the retrieval precision. However, the feature characteristics and the algorithmic procedures of these methods are dramatically different, which is very challenging for the feature-level fusion. This motivates us to investigate how to fuse the ordered retrieval sets, i.e., the ranks of images, given by multiple retrieval methods, to boost the retrieval precision without sacrificing their scalability. In this paper, we model retrieval ranks as graphs of candidate images and propose a graph-based query specific fusion approach, where multiple graphs are merged and reranked by conducting a link analysis on a fused graph. The retrieval quality of an individual method is measured on-the-fly by assessing the consistency of the top candidates' nearest neighborhoods. Hence, it is capable of adaptively integrating the strengths of the retrieval methods using local or holistic features for different query images. This proposed method does not need any supervision, has few parameters, and is easy to implement. Extensive and thorough experiments have been conducted on four public datasets, i.e., the UKbench, Corel-5K, Holidays and the large-scale San Francisco Landmarks datasets. Our proposed method has achieved very competitive performance, including state-of-the-art results on several data sets, e.g., the N-S score 3.83 for UKbench. compact hashing codes UKbench public datasets image retrieval algorithms query specific rank fusion link analysis graph-based query specific fusion approach Scalability Holidays public datasets image fusion feature-level fusion Visualization Vocabulary multiple retrieval methods top candidate nearest neighborhoods ordered retrieval set fusion Fuses trees (mathematics) Image edge detection vocabulary tree image retrieval Corel-5K public datasets large-scale San Francisco landmarks datasets Image retrieval Methods Ming Yang oth Cour, Timothee oth Kai Yu oth Metaxas, Dimitris N oth Enthalten in IEEE transactions on pattern analysis and machine intelligence Washington, DC : IEEE Computer Soc., 1979 37(2015), 4, Seite 803-815 (DE-627)129616397 (DE-600)244051-9 (DE-576)015114341 0162-8828 nnns volume:37 year:2015 number:4 pages:803-815 http://dx.doi.org/10.1109/TPAMI.2014.2346201 Volltext http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6873347 http://www.ncbi.nlm.nih.gov/pubmed/26353295 http://search.proquest.com/docview/1663517075 GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-MAT GBV_ILN_30 GBV_ILN_70 GBV_ILN_2014 GBV_ILN_2410 SQ 1100: 54.74 AVZ AR 37 2015 4 803-815 |
| allfieldsSound |
10.1109/TPAMI.2014.2346201 doi PQ20160617 (DE-627)OLC1968068260 (DE-599)GBVOLC1968068260 (PRQ)c2358-e0de399188db465f4c60726f613e212a686433c749aca598af354b16c83666070 (KEY)0091660920150000037000400803queryspecificrankfusionforimageretrieval DE-627 ger DE-627 rakwb eng 620 004 DNB SQ 1100: AVZ rvk 54.74 bkl Shaoting Zhang verfasserin aut Query Specific Rank Fusion for Image Retrieval 2015 Text txt rdacontent ohne Hilfsmittel zu benutzen n rdamedia Band nc rdacarrier Recently two lines of image retrieval algorithms demonstrate excellent scalability: 1) local features indexed by a vocabulary tree, and 2) holistic features indexed by compact hashing codes. Although both of them are able to search visually similar images effectively, their retrieval precision may vary dramatically among queries. Therefore, combining these two types of methods is expected to further enhance the retrieval precision. However, the feature characteristics and the algorithmic procedures of these methods are dramatically different, which is very challenging for the feature-level fusion. This motivates us to investigate how to fuse the ordered retrieval sets, i.e., the ranks of images, given by multiple retrieval methods, to boost the retrieval precision without sacrificing their scalability. In this paper, we model retrieval ranks as graphs of candidate images and propose a graph-based query specific fusion approach, where multiple graphs are merged and reranked by conducting a link analysis on a fused graph. The retrieval quality of an individual method is measured on-the-fly by assessing the consistency of the top candidates' nearest neighborhoods. Hence, it is capable of adaptively integrating the strengths of the retrieval methods using local or holistic features for different query images. This proposed method does not need any supervision, has few parameters, and is easy to implement. Extensive and thorough experiments have been conducted on four public datasets, i.e., the UKbench, Corel-5K, Holidays and the large-scale San Francisco Landmarks datasets. Our proposed method has achieved very competitive performance, including state-of-the-art results on several data sets, e.g., the N-S score 3.83 for UKbench. compact hashing codes UKbench public datasets image retrieval algorithms query specific rank fusion link analysis graph-based query specific fusion approach Scalability Holidays public datasets image fusion feature-level fusion Visualization Vocabulary multiple retrieval methods top candidate nearest neighborhoods ordered retrieval set fusion Fuses trees (mathematics) Image edge detection vocabulary tree image retrieval Corel-5K public datasets large-scale San Francisco landmarks datasets Image retrieval Methods Ming Yang oth Cour, Timothee oth Kai Yu oth Metaxas, Dimitris N oth Enthalten in IEEE transactions on pattern analysis and machine intelligence Washington, DC : IEEE Computer Soc., 1979 37(2015), 4, Seite 803-815 (DE-627)129616397 (DE-600)244051-9 (DE-576)015114341 0162-8828 nnns volume:37 year:2015 number:4 pages:803-815 http://dx.doi.org/10.1109/TPAMI.2014.2346201 Volltext http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6873347 http://www.ncbi.nlm.nih.gov/pubmed/26353295 http://search.proquest.com/docview/1663517075 GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-MAT GBV_ILN_30 GBV_ILN_70 GBV_ILN_2014 GBV_ILN_2410 SQ 1100: 54.74 AVZ AR 37 2015 4 803-815 |
| language |
English |
| source |
Enthalten in IEEE transactions on pattern analysis and machine intelligence 37(2015), 4, Seite 803-815 volume:37 year:2015 number:4 pages:803-815 |
| sourceStr |
Enthalten in IEEE transactions on pattern analysis and machine intelligence 37(2015), 4, Seite 803-815 volume:37 year:2015 number:4 pages:803-815 |
| format_phy_str_mv |
Article |
| institution |
findex.gbv.de |
| topic_facet |
compact hashing codes UKbench public datasets image retrieval algorithms query specific rank fusion link analysis graph-based query specific fusion approach Scalability Holidays public datasets image fusion feature-level fusion Visualization Vocabulary multiple retrieval methods top candidate nearest neighborhoods ordered retrieval set fusion Fuses trees (mathematics) Image edge detection vocabulary tree image retrieval Corel-5K public datasets large-scale San Francisco landmarks datasets Image retrieval Methods |
| dewey-raw |
620 |
| isfreeaccess_bool |
false |
| container_title |
IEEE transactions on pattern analysis and machine intelligence |
| authorswithroles_txt_mv |
Shaoting Zhang @@aut@@ Ming Yang @@oth@@ Cour, Timothee @@oth@@ Kai Yu @@oth@@ Metaxas, Dimitris N @@oth@@ |
| publishDateDaySort_date |
2015-01-01T00:00:00Z |
| hierarchy_top_id |
129616397 |
| dewey-sort |
3620 |
| id |
OLC1968068260 |
| language_de |
englisch |
| fullrecord |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a2200265 4500</leader><controlfield tag="001">OLC1968068260</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20220220222517.0</controlfield><controlfield tag="007">tu</controlfield><controlfield tag="008">160206s2015 xx ||||| 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1109/TPAMI.2014.2346201</subfield><subfield code="2">doi</subfield></datafield><datafield tag="028" ind1="5" ind2="2"><subfield code="a">PQ20160617</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)OLC1968068260</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-599)GBVOLC1968068260</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(PRQ)c2358-e0de399188db465f4c60726f613e212a686433c749aca598af354b16c83666070</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(KEY)0091660920150000037000400803queryspecificrankfusionforimageretrieval</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">620</subfield><subfield code="a">004</subfield><subfield code="q">DNB</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">SQ 1100:</subfield><subfield code="q">AVZ</subfield><subfield code="2">rvk</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">54.74</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="100" ind1="0" ind2=" "><subfield code="a">Shaoting Zhang</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Query Specific Rank Fusion for Image Retrieval</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2015</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">Text</subfield><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">ohne Hilfsmittel zu benutzen</subfield><subfield code="b">n</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Band</subfield><subfield code="b">nc</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Recently two lines of image retrieval algorithms demonstrate excellent scalability: 1) local features indexed by a vocabulary tree, and 2) holistic features indexed by compact hashing codes. Although both of them are able to search visually similar images effectively, their retrieval precision may vary dramatically among queries. Therefore, combining these two types of methods is expected to further enhance the retrieval precision. However, the feature characteristics and the algorithmic procedures of these methods are dramatically different, which is very challenging for the feature-level fusion. This motivates us to investigate how to fuse the ordered retrieval sets, i.e., the ranks of images, given by multiple retrieval methods, to boost the retrieval precision without sacrificing their scalability. In this paper, we model retrieval ranks as graphs of candidate images and propose a graph-based query specific fusion approach, where multiple graphs are merged and reranked by conducting a link analysis on a fused graph. The retrieval quality of an individual method is measured on-the-fly by assessing the consistency of the top candidates' nearest neighborhoods. Hence, it is capable of adaptively integrating the strengths of the retrieval methods using local or holistic features for different query images. This proposed method does not need any supervision, has few parameters, and is easy to implement. Extensive and thorough experiments have been conducted on four public datasets, i.e., the UKbench, Corel-5K, Holidays and the large-scale San Francisco Landmarks datasets. Our proposed method has achieved very competitive performance, including state-of-the-art results on several data sets, e.g., the N-S score 3.83 for UKbench.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">compact hashing codes</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">UKbench public datasets</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">image retrieval algorithms</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">query specific rank fusion</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">link analysis</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">graph-based query specific fusion approach</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Scalability</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Holidays public datasets</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">image fusion</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">feature-level fusion</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Visualization</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Vocabulary</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">multiple retrieval methods</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">top candidate nearest neighborhoods</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">ordered retrieval set fusion</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Fuses</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">trees (mathematics)</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Image edge detection</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">vocabulary tree</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">image retrieval</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Corel-5K public datasets</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">large-scale San Francisco landmarks datasets</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Image retrieval</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Methods</subfield></datafield><datafield tag="700" ind1="0" ind2=" "><subfield code="a">Ming Yang</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Cour, Timothee</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="0" ind2=" "><subfield code="a">Kai Yu</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Metaxas, Dimitris N</subfield><subfield code="4">oth</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="t">IEEE transactions on pattern analysis and machine intelligence</subfield><subfield code="d">Washington, DC : IEEE Computer Soc., 1979</subfield><subfield code="g">37(2015), 4, Seite 803-815</subfield><subfield code="w">(DE-627)129616397</subfield><subfield code="w">(DE-600)244051-9</subfield><subfield code="w">(DE-576)015114341</subfield><subfield code="x">0162-8828</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:37</subfield><subfield code="g">year:2015</subfield><subfield code="g">number:4</subfield><subfield code="g">pages:803-815</subfield></datafield><datafield tag="856" ind1="4" ind2="1"><subfield code="u">http://dx.doi.org/10.1109/TPAMI.2014.2346201</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="856" ind1="4" ind2="2"><subfield code="u">http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6873347</subfield></datafield><datafield tag="856" ind1="4" ind2="2"><subfield code="u">http://www.ncbi.nlm.nih.gov/pubmed/26353295</subfield></datafield><datafield tag="856" ind1="4" ind2="2"><subfield code="u">http://search.proquest.com/docview/1663517075</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_OLC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-TEC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-MAT</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_30</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_70</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2014</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2410</subfield></datafield><datafield tag="936" ind1="r" ind2="v"><subfield code="a">SQ 1100:</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">54.74</subfield><subfield code="q">AVZ</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">37</subfield><subfield code="j">2015</subfield><subfield code="e">4</subfield><subfield code="h">803-815</subfield></datafield></record></collection>
|
| author |
Shaoting Zhang |
| spellingShingle |
Shaoting Zhang ddc 620 rvk SQ 1100: bkl 54.74 misc compact hashing codes misc UKbench public datasets misc image retrieval algorithms misc query specific rank fusion misc link analysis misc graph-based query specific fusion approach misc Scalability misc Holidays public datasets misc image fusion misc feature-level fusion misc Visualization misc Vocabulary misc multiple retrieval methods misc top candidate nearest neighborhoods misc ordered retrieval set fusion misc Fuses misc trees (mathematics) misc Image edge detection misc vocabulary tree misc image retrieval misc Corel-5K public datasets misc large-scale San Francisco landmarks datasets misc Image retrieval misc Methods Query Specific Rank Fusion for Image Retrieval |
| authorStr |
Shaoting Zhang |
| ppnlink_with_tag_str_mv |
@@773@@(DE-627)129616397 |
| format |
Article |
| dewey-ones |
620 - Engineering & allied operations 004 - Data processing & computer science |
| delete_txt_mv |
keep |
| author_role |
aut |
| collection |
OLC |
| remote_str |
false |
| illustrated |
Not Illustrated |
| issn |
0162-8828 |
| topic_title |
620 004 DNB SQ 1100: AVZ rvk 54.74 bkl Query Specific Rank Fusion for Image Retrieval compact hashing codes UKbench public datasets image retrieval algorithms query specific rank fusion link analysis graph-based query specific fusion approach Scalability Holidays public datasets image fusion feature-level fusion Visualization Vocabulary multiple retrieval methods top candidate nearest neighborhoods ordered retrieval set fusion Fuses trees (mathematics) Image edge detection vocabulary tree image retrieval Corel-5K public datasets large-scale San Francisco landmarks datasets Image retrieval Methods |
| topic |
ddc 620 rvk SQ 1100: bkl 54.74 misc compact hashing codes misc UKbench public datasets misc image retrieval algorithms misc query specific rank fusion misc link analysis misc graph-based query specific fusion approach misc Scalability misc Holidays public datasets misc image fusion misc feature-level fusion misc Visualization misc Vocabulary misc multiple retrieval methods misc top candidate nearest neighborhoods misc ordered retrieval set fusion misc Fuses misc trees (mathematics) misc Image edge detection misc vocabulary tree misc image retrieval misc Corel-5K public datasets misc large-scale San Francisco landmarks datasets misc Image retrieval misc Methods |
| topic_unstemmed |
ddc 620 rvk SQ 1100: bkl 54.74 misc compact hashing codes misc UKbench public datasets misc image retrieval algorithms misc query specific rank fusion misc link analysis misc graph-based query specific fusion approach misc Scalability misc Holidays public datasets misc image fusion misc feature-level fusion misc Visualization misc Vocabulary misc multiple retrieval methods misc top candidate nearest neighborhoods misc ordered retrieval set fusion misc Fuses misc trees (mathematics) misc Image edge detection misc vocabulary tree misc image retrieval misc Corel-5K public datasets misc large-scale San Francisco landmarks datasets misc Image retrieval misc Methods |
| topic_browse |
ddc 620 rvk SQ 1100: bkl 54.74 misc compact hashing codes misc UKbench public datasets misc image retrieval algorithms misc query specific rank fusion misc link analysis misc graph-based query specific fusion approach misc Scalability misc Holidays public datasets misc image fusion misc feature-level fusion misc Visualization misc Vocabulary misc multiple retrieval methods misc top candidate nearest neighborhoods misc ordered retrieval set fusion misc Fuses misc trees (mathematics) misc Image edge detection misc vocabulary tree misc image retrieval misc Corel-5K public datasets misc large-scale San Francisco landmarks datasets misc Image retrieval misc Methods |
| format_facet |
Aufsätze Gedruckte Aufsätze |
| format_main_str_mv |
Text Zeitschrift/Artikel |
| carriertype_str_mv |
nc |
| author2_variant |
m y my t c tc k y ky d n m dn dnm |
| hierarchy_parent_title |
IEEE transactions on pattern analysis and machine intelligence |
| hierarchy_parent_id |
129616397 |
| dewey-tens |
620 - Engineering 000 - Computer science, knowledge & systems |
| hierarchy_top_title |
IEEE transactions on pattern analysis and machine intelligence |
| isfreeaccess_txt |
false |
| familylinks_str_mv |
(DE-627)129616397 (DE-600)244051-9 (DE-576)015114341 |
| title |
Query Specific Rank Fusion for Image Retrieval |
| ctrlnum |
(DE-627)OLC1968068260 (DE-599)GBVOLC1968068260 (PRQ)c2358-e0de399188db465f4c60726f613e212a686433c749aca598af354b16c83666070 (KEY)0091660920150000037000400803queryspecificrankfusionforimageretrieval |
| title_full |
Query Specific Rank Fusion for Image Retrieval |
| author_sort |
Shaoting Zhang |
| journal |
IEEE transactions on pattern analysis and machine intelligence |
| journalStr |
IEEE transactions on pattern analysis and machine intelligence |
| lang_code |
eng |
| isOA_bool |
false |
| dewey-hundreds |
600 - Technology 000 - Computer science, information & general works |
| recordtype |
marc |
| publishDateSort |
2015 |
| contenttype_str_mv |
txt |
| container_start_page |
803 |
| author_browse |
Shaoting Zhang |
| container_volume |
37 |
| class |
620 004 DNB SQ 1100: AVZ rvk 54.74 bkl |
| format_se |
Aufsätze |
| author-letter |
Shaoting Zhang |
| doi_str_mv |
10.1109/TPAMI.2014.2346201 |
| dewey-full |
620 004 |
| title_sort |
query specific rank fusion for image retrieval |
| title_auth |
Query Specific Rank Fusion for Image Retrieval |
| abstract |
Recently two lines of image retrieval algorithms demonstrate excellent scalability: 1) local features indexed by a vocabulary tree, and 2) holistic features indexed by compact hashing codes. Although both of them are able to search visually similar images effectively, their retrieval precision may vary dramatically among queries. Therefore, combining these two types of methods is expected to further enhance the retrieval precision. However, the feature characteristics and the algorithmic procedures of these methods are dramatically different, which is very challenging for the feature-level fusion. This motivates us to investigate how to fuse the ordered retrieval sets, i.e., the ranks of images, given by multiple retrieval methods, to boost the retrieval precision without sacrificing their scalability. In this paper, we model retrieval ranks as graphs of candidate images and propose a graph-based query specific fusion approach, where multiple graphs are merged and reranked by conducting a link analysis on a fused graph. The retrieval quality of an individual method is measured on-the-fly by assessing the consistency of the top candidates' nearest neighborhoods. Hence, it is capable of adaptively integrating the strengths of the retrieval methods using local or holistic features for different query images. This proposed method does not need any supervision, has few parameters, and is easy to implement. Extensive and thorough experiments have been conducted on four public datasets, i.e., the UKbench, Corel-5K, Holidays and the large-scale San Francisco Landmarks datasets. Our proposed method has achieved very competitive performance, including state-of-the-art results on several data sets, e.g., the N-S score 3.83 for UKbench. |
| abstractGer |
Recently two lines of image retrieval algorithms demonstrate excellent scalability: 1) local features indexed by a vocabulary tree, and 2) holistic features indexed by compact hashing codes. Although both of them are able to search visually similar images effectively, their retrieval precision may vary dramatically among queries. Therefore, combining these two types of methods is expected to further enhance the retrieval precision. However, the feature characteristics and the algorithmic procedures of these methods are dramatically different, which is very challenging for the feature-level fusion. This motivates us to investigate how to fuse the ordered retrieval sets, i.e., the ranks of images, given by multiple retrieval methods, to boost the retrieval precision without sacrificing their scalability. In this paper, we model retrieval ranks as graphs of candidate images and propose a graph-based query specific fusion approach, where multiple graphs are merged and reranked by conducting a link analysis on a fused graph. The retrieval quality of an individual method is measured on-the-fly by assessing the consistency of the top candidates' nearest neighborhoods. Hence, it is capable of adaptively integrating the strengths of the retrieval methods using local or holistic features for different query images. This proposed method does not need any supervision, has few parameters, and is easy to implement. Extensive and thorough experiments have been conducted on four public datasets, i.e., the UKbench, Corel-5K, Holidays and the large-scale San Francisco Landmarks datasets. Our proposed method has achieved very competitive performance, including state-of-the-art results on several data sets, e.g., the N-S score 3.83 for UKbench. |
| abstract_unstemmed |
Recently two lines of image retrieval algorithms demonstrate excellent scalability: 1) local features indexed by a vocabulary tree, and 2) holistic features indexed by compact hashing codes. Although both of them are able to search visually similar images effectively, their retrieval precision may vary dramatically among queries. Therefore, combining these two types of methods is expected to further enhance the retrieval precision. However, the feature characteristics and the algorithmic procedures of these methods are dramatically different, which is very challenging for the feature-level fusion. This motivates us to investigate how to fuse the ordered retrieval sets, i.e., the ranks of images, given by multiple retrieval methods, to boost the retrieval precision without sacrificing their scalability. In this paper, we model retrieval ranks as graphs of candidate images and propose a graph-based query specific fusion approach, where multiple graphs are merged and reranked by conducting a link analysis on a fused graph. The retrieval quality of an individual method is measured on-the-fly by assessing the consistency of the top candidates' nearest neighborhoods. Hence, it is capable of adaptively integrating the strengths of the retrieval methods using local or holistic features for different query images. This proposed method does not need any supervision, has few parameters, and is easy to implement. Extensive and thorough experiments have been conducted on four public datasets, i.e., the UKbench, Corel-5K, Holidays and the large-scale San Francisco Landmarks datasets. Our proposed method has achieved very competitive performance, including state-of-the-art results on several data sets, e.g., the N-S score 3.83 for UKbench. |
| collection_details |
GBV_USEFLAG_A SYSFLAG_A GBV_OLC SSG-OLC-TEC SSG-OLC-MAT GBV_ILN_30 GBV_ILN_70 GBV_ILN_2014 GBV_ILN_2410 |
| container_issue |
4 |
| title_short |
Query Specific Rank Fusion for Image Retrieval |
| url |
http://dx.doi.org/10.1109/TPAMI.2014.2346201 http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6873347 http://www.ncbi.nlm.nih.gov/pubmed/26353295 http://search.proquest.com/docview/1663517075 |
| remote_bool |
false |
| author2 |
Ming Yang Cour, Timothee Kai Yu Metaxas, Dimitris N |
| author2Str |
Ming Yang Cour, Timothee Kai Yu Metaxas, Dimitris N |
| ppnlink |
129616397 |
| mediatype_str_mv |
n |
| isOA_txt |
false |
| hochschulschrift_bool |
false |
| author2_role |
oth oth oth oth |
| doi_str |
10.1109/TPAMI.2014.2346201 |
| up_date |
2024-07-04T02:33:56.367Z |
| _version_ |
1803614092128157696 |
| fullrecord_marcxml |
<?xml version="1.0" encoding="UTF-8"?><collection xmlns="http://www.loc.gov/MARC21/slim"><record><leader>01000caa a2200265 4500</leader><controlfield tag="001">OLC1968068260</controlfield><controlfield tag="003">DE-627</controlfield><controlfield tag="005">20220220222517.0</controlfield><controlfield tag="007">tu</controlfield><controlfield tag="008">160206s2015 xx ||||| 00| ||eng c</controlfield><datafield tag="024" ind1="7" ind2=" "><subfield code="a">10.1109/TPAMI.2014.2346201</subfield><subfield code="2">doi</subfield></datafield><datafield tag="028" ind1="5" ind2="2"><subfield code="a">PQ20160617</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-627)OLC1968068260</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(DE-599)GBVOLC1968068260</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(PRQ)c2358-e0de399188db465f4c60726f613e212a686433c749aca598af354b16c83666070</subfield></datafield><datafield tag="035" ind1=" " ind2=" "><subfield code="a">(KEY)0091660920150000037000400803queryspecificrankfusionforimageretrieval</subfield></datafield><datafield tag="040" ind1=" " ind2=" "><subfield code="a">DE-627</subfield><subfield code="b">ger</subfield><subfield code="c">DE-627</subfield><subfield code="e">rakwb</subfield></datafield><datafield tag="041" ind1=" " ind2=" "><subfield code="a">eng</subfield></datafield><datafield tag="082" ind1="0" ind2="4"><subfield code="a">620</subfield><subfield code="a">004</subfield><subfield code="q">DNB</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">SQ 1100:</subfield><subfield code="q">AVZ</subfield><subfield code="2">rvk</subfield></datafield><datafield tag="084" ind1=" " ind2=" "><subfield code="a">54.74</subfield><subfield code="2">bkl</subfield></datafield><datafield tag="100" ind1="0" ind2=" "><subfield code="a">Shaoting Zhang</subfield><subfield code="e">verfasserin</subfield><subfield code="4">aut</subfield></datafield><datafield tag="245" ind1="1" ind2="0"><subfield code="a">Query Specific Rank Fusion for Image Retrieval</subfield></datafield><datafield tag="264" ind1=" " ind2="1"><subfield code="c">2015</subfield></datafield><datafield tag="336" ind1=" " ind2=" "><subfield code="a">Text</subfield><subfield code="b">txt</subfield><subfield code="2">rdacontent</subfield></datafield><datafield tag="337" ind1=" " ind2=" "><subfield code="a">ohne Hilfsmittel zu benutzen</subfield><subfield code="b">n</subfield><subfield code="2">rdamedia</subfield></datafield><datafield tag="338" ind1=" " ind2=" "><subfield code="a">Band</subfield><subfield code="b">nc</subfield><subfield code="2">rdacarrier</subfield></datafield><datafield tag="520" ind1=" " ind2=" "><subfield code="a">Recently two lines of image retrieval algorithms demonstrate excellent scalability: 1) local features indexed by a vocabulary tree, and 2) holistic features indexed by compact hashing codes. Although both of them are able to search visually similar images effectively, their retrieval precision may vary dramatically among queries. Therefore, combining these two types of methods is expected to further enhance the retrieval precision. However, the feature characteristics and the algorithmic procedures of these methods are dramatically different, which is very challenging for the feature-level fusion. This motivates us to investigate how to fuse the ordered retrieval sets, i.e., the ranks of images, given by multiple retrieval methods, to boost the retrieval precision without sacrificing their scalability. In this paper, we model retrieval ranks as graphs of candidate images and propose a graph-based query specific fusion approach, where multiple graphs are merged and reranked by conducting a link analysis on a fused graph. The retrieval quality of an individual method is measured on-the-fly by assessing the consistency of the top candidates' nearest neighborhoods. Hence, it is capable of adaptively integrating the strengths of the retrieval methods using local or holistic features for different query images. This proposed method does not need any supervision, has few parameters, and is easy to implement. Extensive and thorough experiments have been conducted on four public datasets, i.e., the UKbench, Corel-5K, Holidays and the large-scale San Francisco Landmarks datasets. Our proposed method has achieved very competitive performance, including state-of-the-art results on several data sets, e.g., the N-S score 3.83 for UKbench.</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">compact hashing codes</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">UKbench public datasets</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">image retrieval algorithms</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">query specific rank fusion</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">link analysis</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">graph-based query specific fusion approach</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Scalability</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Holidays public datasets</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">image fusion</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">feature-level fusion</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Visualization</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Vocabulary</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">multiple retrieval methods</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">top candidate nearest neighborhoods</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">ordered retrieval set fusion</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Fuses</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">trees (mathematics)</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Image edge detection</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">vocabulary tree</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">image retrieval</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Corel-5K public datasets</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">large-scale San Francisco landmarks datasets</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Image retrieval</subfield></datafield><datafield tag="650" ind1=" " ind2="4"><subfield code="a">Methods</subfield></datafield><datafield tag="700" ind1="0" ind2=" "><subfield code="a">Ming Yang</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Cour, Timothee</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="0" ind2=" "><subfield code="a">Kai Yu</subfield><subfield code="4">oth</subfield></datafield><datafield tag="700" ind1="1" ind2=" "><subfield code="a">Metaxas, Dimitris N</subfield><subfield code="4">oth</subfield></datafield><datafield tag="773" ind1="0" ind2="8"><subfield code="i">Enthalten in</subfield><subfield code="t">IEEE transactions on pattern analysis and machine intelligence</subfield><subfield code="d">Washington, DC : IEEE Computer Soc., 1979</subfield><subfield code="g">37(2015), 4, Seite 803-815</subfield><subfield code="w">(DE-627)129616397</subfield><subfield code="w">(DE-600)244051-9</subfield><subfield code="w">(DE-576)015114341</subfield><subfield code="x">0162-8828</subfield><subfield code="7">nnns</subfield></datafield><datafield tag="773" ind1="1" ind2="8"><subfield code="g">volume:37</subfield><subfield code="g">year:2015</subfield><subfield code="g">number:4</subfield><subfield code="g">pages:803-815</subfield></datafield><datafield tag="856" ind1="4" ind2="1"><subfield code="u">http://dx.doi.org/10.1109/TPAMI.2014.2346201</subfield><subfield code="3">Volltext</subfield></datafield><datafield tag="856" ind1="4" ind2="2"><subfield code="u">http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=6873347</subfield></datafield><datafield tag="856" ind1="4" ind2="2"><subfield code="u">http://www.ncbi.nlm.nih.gov/pubmed/26353295</subfield></datafield><datafield tag="856" ind1="4" ind2="2"><subfield code="u">http://search.proquest.com/docview/1663517075</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_USEFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SYSFLAG_A</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_OLC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-TEC</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">SSG-OLC-MAT</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_30</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_70</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2014</subfield></datafield><datafield tag="912" ind1=" " ind2=" "><subfield code="a">GBV_ILN_2410</subfield></datafield><datafield tag="936" ind1="r" ind2="v"><subfield code="a">SQ 1100:</subfield></datafield><datafield tag="936" ind1="b" ind2="k"><subfield code="a">54.74</subfield><subfield code="q">AVZ</subfield></datafield><datafield tag="951" ind1=" " ind2=" "><subfield code="a">AR</subfield></datafield><datafield tag="952" ind1=" " ind2=" "><subfield code="d">37</subfield><subfield code="j">2015</subfield><subfield code="e">4</subfield><subfield code="h">803-815</subfield></datafield></record></collection>
|
| score |
7.4010096 |