article · DSpace - NTUA (National Technical University of Athens) · 2010

Feature map hashing

Yannis Avrithis, Giorgos Tolias, Yannis Kalantidis

Cited by 22Open access

Abstract

We present a new approach to image indexing and retrieval, which integrates appearance with global image geometry in the indexing process, while enjoying robustness against viewpoint change, photometric variations, occlusion, and background clutter. We exploit shape parameters of local features to estimate image alignment via a single correspondence. Then, for each feature, we construct a sparse spatial map of all remaining features, encoding their normalized position and appearance, typically vector quantized to visual word. An image is represented by a collection of such feature maps and RANSAC-like matching is reduced to a number of set intersections. Because the induced dissimilarity is still not a metric, we extend min-wise independent permutations to collections of sets and derive a similarity measure for feature map collections. We then exploit sparseness to build an inverted file whereby the retrieval process is sub-linear in the total number of images, ideally linear in the number of relevant ones. We achieve excellent performance on 10^4 images, with a query time in the order of milliseconds.

Cite this paper

APA 7

Avrithis, Y., Tolias, G., & Kalantidis, Y. (2010). Feature map hashing. DSpace - NTUA (National Technical University of Athens), 231–240. https://doi.org/10.1145/1873951.1873985

MLA 9

Avrithis, Yannis, et al. “Feature map hashing.” DSpace - NTUA (National Technical University of Athens), 2010, pp. 231–40, https://doi.org/10.1145/1873951.1873985.

Chicago

Avrithis, Yannis, Giorgos Tolias, and Yannis Kalantidis. 2010. “Feature map hashing.” DSpace - NTUA (National Technical University of Athens): 231–40. https://doi.org/10.1145/1873951.1873985.

Harvard

Avrithis, Y., Tolias, G. and Kalantidis, Y. (2010) ‘Feature map hashing’, DSpace - NTUA (National Technical University of Athens), pp. 231–240. Available at: https://doi.org/10.1145/1873951.1873985.

BibTeX

@article{avrithis2010feature,
  title = {{Feature map hashing}},
  author = {Avrithis, Yannis and Tolias, Giorgos and Kalantidis, Yannis},
  journal = {DSpace - NTUA (National Technical University of Athens)},
  pages = {231--240},
  year = {2010},
  doi = {10.1145/1873951.1873985}
}

Made from the index record. Check names, capitalisation and page numbers against the paper. Need another style? AI for Research has over 10,000.