The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
PCA-SIFT는 그라디언트 이미지 패치에 PCA를 적용하여 SIFT의 고차원성(128차원)을 줄이는 것을 목표로 하는 SIFT의 확장입니다. 그러나 PCA는 전역 특성 특성과 비지도 알고리즘으로 인해 인식에 대한 차별적 표현이 아닙니다. 또한 PCA, ICA와 같은 선형 방법은 비선형성의 경우 실패할 수 있습니다. 본 논문에서는 다음과 같은 새로운 판별 방법을 제안한다. 감독 감독된 ICA 기반 로컬 이미지 설명자와 결합된 비선형 커널 접근 방식을 사용하는 커널 ICA(SKICA). 우리의 접근 방식은 지도 학습의 장점과 커널의 비선형 속성을 혼합합니다. 5개의 서로 다른 테스트 데이터 세트를 사용하여 SKICA 설명자가 동일한 차원을 가진 다른 관련 접근 방식보다 더 나은 객체 인식 성능을 생성한다는 것을 보여줍니다. SKICA 기반 표현은 로컬 감도, 비선형 독립성 및 고급 분리성을 갖추고 있어 로컬 이미지 설명자에 대한 효과적인 방법을 제공합니다.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
부
Masaki YAMAZAKI, Sidney FELS, "Local Image Descriptors Using Supervised Kernel ICA" in IEICE TRANSACTIONS on Information,
vol. E92-D, no. 9, pp. 1745-1751, September 2009, doi: 10.1587/transinf.E92.D.1745.
Abstract: PCA-SIFT is an extension to SIFT which aims to reduce SIFT's high dimensionality (128 dimensions) by applying PCA to the gradient image patches. However PCA is not a discriminative representation for recognition due to its global feature nature and unsupervised algorithm. In addition, linear methods such as PCA and ICA can fail in the case of non-linearity. In this paper, we propose a new discriminative method called Supervised Kernel ICA (SKICA) that uses a non-linear kernel approach combined with Supervised ICA-based local image descriptors. Our approach blends the advantages of supervised learning with nonlinear properties of kernels. Using five different test data sets we show that the SKICA descriptors produce better object recognition performance than other related approaches with the same dimensionality. The SKICA-based representation has local sensitivity, non-linear independence and high class separability providing an effective method for local image descriptors.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.E92.D.1745/_p
부
@ARTICLE{e92-d_9_1745,
author={Masaki YAMAZAKI, Sidney FELS, },
journal={IEICE TRANSACTIONS on Information},
title={Local Image Descriptors Using Supervised Kernel ICA},
year={2009},
volume={E92-D},
number={9},
pages={1745-1751},
abstract={PCA-SIFT is an extension to SIFT which aims to reduce SIFT's high dimensionality (128 dimensions) by applying PCA to the gradient image patches. However PCA is not a discriminative representation for recognition due to its global feature nature and unsupervised algorithm. In addition, linear methods such as PCA and ICA can fail in the case of non-linearity. In this paper, we propose a new discriminative method called Supervised Kernel ICA (SKICA) that uses a non-linear kernel approach combined with Supervised ICA-based local image descriptors. Our approach blends the advantages of supervised learning with nonlinear properties of kernels. Using five different test data sets we show that the SKICA descriptors produce better object recognition performance than other related approaches with the same dimensionality. The SKICA-based representation has local sensitivity, non-linear independence and high class separability providing an effective method for local image descriptors.},
keywords={},
doi={10.1587/transinf.E92.D.1745},
ISSN={1745-1361},
month={September},}
부
TY - JOUR
TI - Local Image Descriptors Using Supervised Kernel ICA
T2 - IEICE TRANSACTIONS on Information
SP - 1745
EP - 1751
AU - Masaki YAMAZAKI
AU - Sidney FELS
PY - 2009
DO - 10.1587/transinf.E92.D.1745
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E92-D
IS - 9
JA - IEICE TRANSACTIONS on Information
Y1 - September 2009
AB - PCA-SIFT is an extension to SIFT which aims to reduce SIFT's high dimensionality (128 dimensions) by applying PCA to the gradient image patches. However PCA is not a discriminative representation for recognition due to its global feature nature and unsupervised algorithm. In addition, linear methods such as PCA and ICA can fail in the case of non-linearity. In this paper, we propose a new discriminative method called Supervised Kernel ICA (SKICA) that uses a non-linear kernel approach combined with Supervised ICA-based local image descriptors. Our approach blends the advantages of supervised learning with nonlinear properties of kernels. Using five different test data sets we show that the SKICA descriptors produce better object recognition performance than other related approaches with the same dimensionality. The SKICA-based representation has local sensitivity, non-linear independence and high class separability providing an effective method for local image descriptors.
ER -