Fast Pose Estimation with Parameter Sensitive Hashing
Example-based methods are effective for parameter estimation problems when the underlying system is simple or the dimensionality of the input is low. For complex and high-dimensional problems such as pose estimation, the number of required examples and the computational complexity rapidly becme...
Main Authors: | , , |
---|---|
Language: | en_US |
Published: |
2004
|
Subjects: | |
Online Access: | http://hdl.handle.net/1721.1/6715 |
_version_ | 1811090074328104960 |
---|---|
author | Shakhnarovich, Gregory Viola, Paul Darrell, Trevor |
author_facet | Shakhnarovich, Gregory Viola, Paul Darrell, Trevor |
author_sort | Shakhnarovich, Gregory |
collection | MIT |
description | Example-based methods are effective for parameter estimation problems when the underlying system is simple or the dimensionality of the input is low. For complex and high-dimensional problems such as pose estimation, the number of required examples and the computational complexity rapidly becme prohibitively high. We introduce a new algorithm that learns a set of hashing functions that efficiently index examples relevant to a particular estimation task. Our algorithm extends a recently developed method for locality-sensitive hashing, which finds approximate neighbors in time sublinear in the number of examples. This method depends critically on the choice of hash functions; we show how to find the set of hash functions that are optimally relevant to a particular estimation problem. Experiments demonstrate that the resulting algorithm, which we call Parameter-Sensitive Hashing, can rapidly and accurately estimate the articulated pose of human figures from a large database of example images. |
first_indexed | 2024-09-23T14:32:27Z |
id | mit-1721.1/6715 |
institution | Massachusetts Institute of Technology |
language | en_US |
last_indexed | 2024-09-23T14:32:27Z |
publishDate | 2004 |
record_format | dspace |
spelling | mit-1721.1/67152019-04-10T16:35:42Z Fast Pose Estimation with Parameter Sensitive Hashing Shakhnarovich, Gregory Viola, Paul Darrell, Trevor AI parameter estimation nearest neighbor locally weighted learning Example-based methods are effective for parameter estimation problems when the underlying system is simple or the dimensionality of the input is low. For complex and high-dimensional problems such as pose estimation, the number of required examples and the computational complexity rapidly becme prohibitively high. We introduce a new algorithm that learns a set of hashing functions that efficiently index examples relevant to a particular estimation task. Our algorithm extends a recently developed method for locality-sensitive hashing, which finds approximate neighbors in time sublinear in the number of examples. This method depends critically on the choice of hash functions; we show how to find the set of hash functions that are optimally relevant to a particular estimation problem. Experiments demonstrate that the resulting algorithm, which we call Parameter-Sensitive Hashing, can rapidly and accurately estimate the articulated pose of human figures from a large database of example images. 2004-10-08T20:38:53Z 2004-10-08T20:38:53Z 2003-04-18 AIM-2003-009 http://hdl.handle.net/1721.1/6715 en_US AIM-2003-009 12 p. 5030222 bytes 6836715 bytes application/postscript application/pdf application/postscript application/pdf |
spellingShingle | AI parameter estimation nearest neighbor locally weighted learning Shakhnarovich, Gregory Viola, Paul Darrell, Trevor Fast Pose Estimation with Parameter Sensitive Hashing |
title | Fast Pose Estimation with Parameter Sensitive Hashing |
title_full | Fast Pose Estimation with Parameter Sensitive Hashing |
title_fullStr | Fast Pose Estimation with Parameter Sensitive Hashing |
title_full_unstemmed | Fast Pose Estimation with Parameter Sensitive Hashing |
title_short | Fast Pose Estimation with Parameter Sensitive Hashing |
title_sort | fast pose estimation with parameter sensitive hashing |
topic | AI parameter estimation nearest neighbor locally weighted learning |
url | http://hdl.handle.net/1721.1/6715 |
work_keys_str_mv | AT shakhnarovichgregory fastposeestimationwithparametersensitivehashing AT violapaul fastposeestimationwithparametersensitivehashing AT darrelltrevor fastposeestimationwithparametersensitivehashing |