Randomized sub-vectors hashing for high-dimensional image feature matching

Heng Yang; Qing Wang; Zhoucan He

doi:10.1145/1459359.1459465

Randomized sub-vectors hashing for high-dimensional image feature matching

Heng Yang, Qing Wang, Zhoucan He

School of Computer Science

Northwestern Polytechnical University Xian

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

5 Scopus citations

Abstract

High-dimensional image feature matching is an important part of many image matching based problems in computer vision which are solved by local invariant features. In this paper, we propose a new indexing/searching method based on Randomized Sub-Vectors Hashing (called RSVH) for high-dimensional image feature matching. The essential of the proposed idea is that the feature vectors are considered similar (measured by Euclidean distance) when the L2 norms of their corresponding randomized sub-vectors are approximately same respectively. Experimental results have demonstrated that our algorithm can perform much better than the famous BBF (Best-Bin-First) and LSH (Locality Sensitive Hashing) algorithms in extensive image matching and image retrieval applications.

Original language	English
Title of host publication	MM'08 - Proceedings of the 2008 ACM International Conference on Multimedia, with co-located Symposium and Workshops
Pages	705-708
Number of pages	4
DOIs	https://doi.org/10.1145/1459359.1459465
State	Published - 2008
Event	16th ACM International Conference on Multimedia, MM '08 - Vancouver, BC, Canada Duration: 26 Oct 2008 → 31 Oct 2008

Publication series

Name	MM'08 - Proceedings of the 2008 ACM International Conference on Multimedia, with co-located Symposium and Workshops

Conference

Conference	16th ACM International Conference on Multimedia, MM '08
Country/Territory	Canada
City	Vancouver, BC
Period	26/10/08 → 31/10/08

Keywords

High-dimensional feature matching
Nearest neighbor searching
Randomized sub-vectors hashing

Access to Document

10.1145/1459359.1459465

Cite this

Yang, H., Wang, Q., & He, Z. (2008). Randomized sub-vectors hashing for high-dimensional image feature matching. In MM'08 - Proceedings of the 2008 ACM International Conference on Multimedia, with co-located Symposium and Workshops (pp. 705-708). (MM'08 - Proceedings of the 2008 ACM International Conference on Multimedia, with co-located Symposium and Workshops). https://doi.org/10.1145/1459359.1459465

Yang, Heng ; Wang, Qing ; He, Zhoucan. / Randomized sub-vectors hashing for high-dimensional image feature matching. MM'08 - Proceedings of the 2008 ACM International Conference on Multimedia, with co-located Symposium and Workshops. 2008. pp. 705-708 (MM'08 - Proceedings of the 2008 ACM International Conference on Multimedia, with co-located Symposium and Workshops).

@inproceedings{c8a9e3543a184df69bab67a43e087029,

title = "Randomized sub-vectors hashing for high-dimensional image feature matching",

abstract = "High-dimensional image feature matching is an important part of many image matching based problems in computer vision which are solved by local invariant features. In this paper, we propose a new indexing/searching method based on Randomized Sub-Vectors Hashing (called RSVH) for high-dimensional image feature matching. The essential of the proposed idea is that the feature vectors are considered similar (measured by Euclidean distance) when the L2 norms of their corresponding randomized sub-vectors are approximately same respectively. Experimental results have demonstrated that our algorithm can perform much better than the famous BBF (Best-Bin-First) and LSH (Locality Sensitive Hashing) algorithms in extensive image matching and image retrieval applications.",

keywords = "High-dimensional feature matching, Nearest neighbor searching, Randomized sub-vectors hashing",

author = "Heng Yang and Qing Wang and Zhoucan He",

year = "2008",

doi = "10.1145/1459359.1459465",

language = "英语",

isbn = "9781605583037",

series = "MM'08 - Proceedings of the 2008 ACM International Conference on Multimedia, with co-located Symposium and Workshops",

pages = "705--708",

booktitle = "MM'08 - Proceedings of the 2008 ACM International Conference on Multimedia, with co-located Symposium and Workshops",

note = "16th ACM International Conference on Multimedia, MM '08 ; Conference date: 26-10-2008 Through 31-10-2008",

}

Yang, H, Wang, Q & He, Z 2008, Randomized sub-vectors hashing for high-dimensional image feature matching. in MM'08 - Proceedings of the 2008 ACM International Conference on Multimedia, with co-located Symposium and Workshops. MM'08 - Proceedings of the 2008 ACM International Conference on Multimedia, with co-located Symposium and Workshops, pp. 705-708, 16th ACM International Conference on Multimedia, MM '08, Vancouver, BC, Canada, 26/10/08. https://doi.org/10.1145/1459359.1459465

Randomized sub-vectors hashing for high-dimensional image feature matching. / Yang, Heng; Wang, Qing; He, Zhoucan.
MM'08 - Proceedings of the 2008 ACM International Conference on Multimedia, with co-located Symposium and Workshops. 2008. p. 705-708 (MM'08 - Proceedings of the 2008 ACM International Conference on Multimedia, with co-located Symposium and Workshops).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Randomized sub-vectors hashing for high-dimensional image feature matching

AU - Yang, Heng

AU - Wang, Qing

AU - He, Zhoucan

PY - 2008

Y1 - 2008

N2 - High-dimensional image feature matching is an important part of many image matching based problems in computer vision which are solved by local invariant features. In this paper, we propose a new indexing/searching method based on Randomized Sub-Vectors Hashing (called RSVH) for high-dimensional image feature matching. The essential of the proposed idea is that the feature vectors are considered similar (measured by Euclidean distance) when the L2 norms of their corresponding randomized sub-vectors are approximately same respectively. Experimental results have demonstrated that our algorithm can perform much better than the famous BBF (Best-Bin-First) and LSH (Locality Sensitive Hashing) algorithms in extensive image matching and image retrieval applications.

AB - High-dimensional image feature matching is an important part of many image matching based problems in computer vision which are solved by local invariant features. In this paper, we propose a new indexing/searching method based on Randomized Sub-Vectors Hashing (called RSVH) for high-dimensional image feature matching. The essential of the proposed idea is that the feature vectors are considered similar (measured by Euclidean distance) when the L2 norms of their corresponding randomized sub-vectors are approximately same respectively. Experimental results have demonstrated that our algorithm can perform much better than the famous BBF (Best-Bin-First) and LSH (Locality Sensitive Hashing) algorithms in extensive image matching and image retrieval applications.

KW - High-dimensional feature matching

KW - Nearest neighbor searching

KW - Randomized sub-vectors hashing

UR - http://www.scopus.com/inward/record.url?scp=70350641516&partnerID=8YFLogxK

U2 - 10.1145/1459359.1459465

DO - 10.1145/1459359.1459465

M3 - 会议稿件

AN - SCOPUS:70350641516

SN - 9781605583037

T3 - MM'08 - Proceedings of the 2008 ACM International Conference on Multimedia, with co-located Symposium and Workshops

SP - 705

EP - 708

BT - MM'08 - Proceedings of the 2008 ACM International Conference on Multimedia, with co-located Symposium and Workshops

T2 - 16th ACM International Conference on Multimedia, MM '08

Y2 - 26 October 2008 through 31 October 2008

ER -

Yang H, Wang Q, He Z. Randomized sub-vectors hashing for high-dimensional image feature matching. In MM'08 - Proceedings of the 2008 ACM International Conference on Multimedia, with co-located Symposium and Workshops. 2008. p. 705-708. (MM'08 - Proceedings of the 2008 ACM International Conference on Multimedia, with co-located Symposium and Workshops). doi: 10.1145/1459359.1459465

Randomized sub-vectors hashing for high-dimensional image feature matching

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this