TY - GEN
T1 - Fast subspace search via grassmannian based hashing
AU - Wang, Xu
AU - Atev, Stefan
AU - Wright, John
AU - Lerman, Gilad
N1 - Copyright:
Copyright 2014 Elsevier B.V., All rights reserved.
PY - 2013
Y1 - 2013
N2 - The problem of efficiently deciding which of a database of models is most similar to a given input query arises throughout modern computer vision. Motivated by applications in recognition, image retrieval and optimization, there has been significant recent interest in the variant of this problem in which the database models are linear subspaces and the input is either a point or a subspace. Current approaches to this problem have poor scaling in high dimensions, and may not guarantee sub linear query complexity. We present a new approach to approximate nearest subspace search, based on a simple, new locality sensitive hash for subspaces. Our approach allows point-to-subspace query for a database of subspaces of arbitrary dimension d, in a time that depends sub linearly on the number of subspaces in the database. The query complexity of our algorithm is linear in the ambient dimension D, allowing it to be directly applied to high-dimensional imagery data. Numerical experiments on model problems in image repatching and automatic face recognition confirm the advantages of our algorithm in terms of both speed and accuracy.
AB - The problem of efficiently deciding which of a database of models is most similar to a given input query arises throughout modern computer vision. Motivated by applications in recognition, image retrieval and optimization, there has been significant recent interest in the variant of this problem in which the database models are linear subspaces and the input is either a point or a subspace. Current approaches to this problem have poor scaling in high dimensions, and may not guarantee sub linear query complexity. We present a new approach to approximate nearest subspace search, based on a simple, new locality sensitive hash for subspaces. Our approach allows point-to-subspace query for a database of subspaces of arbitrary dimension d, in a time that depends sub linearly on the number of subspaces in the database. The query complexity of our algorithm is linear in the ambient dimension D, allowing it to be directly applied to high-dimensional imagery data. Numerical experiments on model problems in image repatching and automatic face recognition confirm the advantages of our algorithm in terms of both speed and accuracy.
KW - Grassmannian Based Hashing
KW - Locality Sensitive Hashing
KW - Subspace Search
UR - http://www.scopus.com/inward/record.url?scp=84898822518&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84898822518&partnerID=8YFLogxK
U2 - 10.1109/ICCV.2013.345
DO - 10.1109/ICCV.2013.345
M3 - Conference contribution
AN - SCOPUS:84898822518
SN - 9781479928392
T3 - Proceedings of the IEEE International Conference on Computer Vision
SP - 2776
EP - 2783
BT - Proceedings - 2013 IEEE International Conference on Computer Vision, ICCV 2013
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2013 14th IEEE International Conference on Computer Vision, ICCV 2013
Y2 - 1 December 2013 through 8 December 2013
ER -