Improving Bag-of-Features for Large Scale Image Search. Jégou, H., Douze, M., & Schmid, C. International Journal of Computer Vision, 87(3):316–336, May, 2010. Paper doi abstract bibtex This article improves recent methods for large scale image search. We first analyze the bag-of-features approach in the framework of approximate nearest neighbor search. This leads us to derive a more precise representation based on 1) Hamming embedding (HE) and 2) weak geometric consistency constraints (WGC). HE provides binary signatures that refine the matching based on visual words. WGC filters matching descriptors that are not consistent in terms of angle and scale. HE and WGC are integrated within an inverted file and are efficiently exploited for all images in the dataset. We then introduce a graph-structured quantizer which significantly speeds up the assignment of the descriptors to visual words. A comparison with the state of the art shows the interest of our approach when high accuracy is needed.
@article{jegou_improving_2010,
title = {Improving {Bag}-of-{Features} for {Large} {Scale} {Image} {Search}},
volume = {87},
issn = {0920-5691, 1573-1405},
url = {http://link.springer.com/10.1007/s11263-009-0285-2},
doi = {10.1007/s11263-009-0285-2},
abstract = {This article improves recent methods for large scale image search. We first analyze the bag-of-features approach in the framework of approximate nearest neighbor search. This leads us to derive a more precise representation based on 1) Hamming embedding (HE) and 2) weak geometric consistency constraints (WGC). HE provides binary signatures that refine the matching based on visual words. WGC filters matching descriptors that are not consistent in terms of angle and scale. HE and WGC are integrated within an inverted file and are efficiently exploited for all images in the dataset. We then introduce a graph-structured quantizer which significantly speeds up the assignment of the descriptors to visual words. A comparison with the state of the art shows the interest of our approach when high accuracy is needed.},
language = {en},
number = {3},
urldate = {2022-03-02},
journal = {International Journal of Computer Vision},
author = {Jégou, Hervé and Douze, Matthijs and Schmid, Cordelia},
month = may,
year = {2010},
pages = {316--336},
}
Downloads: 0
{"_id":"9pp6E5deyncARtFz2","bibbaseid":"jgou-douze-schmid-improvingbagoffeaturesforlargescaleimagesearch-2010","author_short":["Jégou, H.","Douze, M.","Schmid, C."],"bibdata":{"bibtype":"article","type":"article","title":"Improving Bag-of-Features for Large Scale Image Search","volume":"87","issn":"0920-5691, 1573-1405","url":"http://link.springer.com/10.1007/s11263-009-0285-2","doi":"10.1007/s11263-009-0285-2","abstract":"This article improves recent methods for large scale image search. We first analyze the bag-of-features approach in the framework of approximate nearest neighbor search. This leads us to derive a more precise representation based on 1) Hamming embedding (HE) and 2) weak geometric consistency constraints (WGC). HE provides binary signatures that refine the matching based on visual words. WGC filters matching descriptors that are not consistent in terms of angle and scale. HE and WGC are integrated within an inverted file and are efficiently exploited for all images in the dataset. We then introduce a graph-structured quantizer which significantly speeds up the assignment of the descriptors to visual words. A comparison with the state of the art shows the interest of our approach when high accuracy is needed.","language":"en","number":"3","urldate":"2022-03-02","journal":"International Journal of Computer Vision","author":[{"propositions":[],"lastnames":["Jégou"],"firstnames":["Hervé"],"suffixes":[]},{"propositions":[],"lastnames":["Douze"],"firstnames":["Matthijs"],"suffixes":[]},{"propositions":[],"lastnames":["Schmid"],"firstnames":["Cordelia"],"suffixes":[]}],"month":"May","year":"2010","pages":"316–336","bibtex":"@article{jegou_improving_2010,\n\ttitle = {Improving {Bag}-of-{Features} for {Large} {Scale} {Image} {Search}},\n\tvolume = {87},\n\tissn = {0920-5691, 1573-1405},\n\turl = {http://link.springer.com/10.1007/s11263-009-0285-2},\n\tdoi = {10.1007/s11263-009-0285-2},\n\tabstract = {This article improves recent methods for large scale image search. We first analyze the bag-of-features approach in the framework of approximate nearest neighbor search. This leads us to derive a more precise representation based on 1) Hamming embedding (HE) and 2) weak geometric consistency constraints (WGC). HE provides binary signatures that refine the matching based on visual words. WGC filters matching descriptors that are not consistent in terms of angle and scale. HE and WGC are integrated within an inverted file and are efficiently exploited for all images in the dataset. We then introduce a graph-structured quantizer which significantly speeds up the assignment of the descriptors to visual words. A comparison with the state of the art shows the interest of our approach when high accuracy is needed.},\n\tlanguage = {en},\n\tnumber = {3},\n\turldate = {2022-03-02},\n\tjournal = {International Journal of Computer Vision},\n\tauthor = {Jégou, Hervé and Douze, Matthijs and Schmid, Cordelia},\n\tmonth = may,\n\tyear = {2010},\n\tpages = {316--336},\n}\n\n","author_short":["Jégou, H.","Douze, M.","Schmid, C."],"key":"jegou_improving_2010","id":"jegou_improving_2010","bibbaseid":"jgou-douze-schmid-improvingbagoffeaturesforlargescaleimagesearch-2010","role":"author","urls":{"Paper":"http://link.springer.com/10.1007/s11263-009-0285-2"},"metadata":{"authorlinks":{}},"html":""},"bibtype":"article","biburl":"https://bibbase.org/zotero/mxmplx","dataSources":["aXmRAq63YsH7a3ufx"],"keywords":[],"search_terms":["improving","bag","features","large","scale","image","search","jégou","douze","schmid"],"title":"Improving Bag-of-Features for Large Scale Image Search","year":2010}