Campus Event Calendar
Event Entry
What and Who
Large Scale Image Search
Dr. Cordelia Schmid
INRIA
Distinguished Speaker Series Max
Planck Institute for Informatics and Cluster of Excellence Multimodal Computing and Interaction
AG 1, AG 2, AG 3, AG 4, AG 5, SWS, RG1, RG2
Public Audience
English
Note: We use this to send email in the morning.
Date, Time and Location
Monday, 22 September 2008
11:00
45 Minutes
E1 4
024
Saarbrücken
Abstract
We address the problem of large scale image search, for which many
recent methods use a bag-of-features image representation. We shows
the sub-optimality of such a representation for matching descriptors
and 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 system
and are efficiently exploited even in the case of very large
datasets. Experiments performed on a dataset of one million of images
show a significant improvement due to the binary signatures and the
weak geometric consistency constraints, as well as their
efficiency. Estimation of the full geometric transformation, i.e., a
re-ranking step on a short list of images, is complementary to our
weak geometric consistency constraints and allows to further improve
the accuracy.
This is joint work with H. Jegou and M. Douze.
recent methods use a bag-of-features image representation. We shows
the sub-optimality of such a representation for matching descriptors
and 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 system
and are efficiently exploited even in the case of very large
datasets. Experiments performed on a dataset of one million of images
show a significant improvement due to the binary signatures and the
weak geometric consistency constraints, as well as their
efficiency. Estimation of the full geometric transformation, i.e., a
re-ranking step on a short list of images, is complementary to our
weak geometric consistency constraints and allows to further improve
the accuracy.
This is joint work with H. Jegou and M. Douze.
Contact
Roxane Wetzel
900
--email hidden
passcode not visible
logged in users only
Roxane Wetzel, 09/10/2008 16:48
Roxane Wetzel, 09/05/2008 13:55
Roxane Wetzel, 07/16/2008 11:08
Roxane Wetzel, 06/16/2008 09:04
Roxane Wetzel, 06/11/2008 14:32
Roxane Wetzel, 05/30/2008 13:32
Roxane Wetzel, 05/30/2008 13:04 -- Created document.
Roxane Wetzel, 09/05/2008 13:55
Roxane Wetzel, 07/16/2008 11:08
Roxane Wetzel, 06/16/2008 09:04
Roxane Wetzel, 06/11/2008 14:32
Roxane Wetzel, 05/30/2008 13:32
Roxane Wetzel, 05/30/2008 13:04 -- Created document.