CompactKdt: Compact signatures for accurate large scale object recognition

Mohamed Aly*, Mario Munich, Pietro Perona

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

3 Scopus citations

Abstract

We present a novel algorithm, Compact Kd-Trees (CompactKdt), that achieves state-of-the-art performance in searching large scale object image collections. The algorithm uses an order of magnitude less storage and computations by making use of both the full local features (e.g. SIFT) and their compact binary signatures to build and search the K-Tree. We compare classical PCA dimensionality reduction to three methods for generating compact binary representations for the features: Spectral Hashing, Locality Sensitive Hashing, and Locality Sensitive Binary Codes. CompactKdt achieves significant performance gain over using the binary signatures alone, and comparable performance to using the full features alone. Finally, our experiments show significantly better performance than the state-of-the-art Bag of Words (BoW) methods with equivalent or less storage and computational cost.

Original languageEnglish (US)
Title of host publication2012 IEEE Workshop on the Applications of Computer Vision, WACV 2012
Pages505-512
Number of pages8
DOIs
StatePublished - 2012
Externally publishedYes
Event2012 IEEE Workshop on the Applications of Computer Vision, WACV 2012 - Breckenridge, CO, United States
Duration: Jan 9 2012Jan 11 2012

Other

Other2012 IEEE Workshop on the Applications of Computer Vision, WACV 2012
CountryUnited States
CityBreckenridge, CO
Period01/9/1201/11/12

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Computer Science Applications

Fingerprint Dive into the research topics of 'CompactKdt: Compact signatures for accurate large scale object recognition'. Together they form a unique fingerprint.

Cite this