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
Recentemente, a técnica de aprendizagem de dicionário multivisualização atraiu muito interesse de pesquisa. Embora vários métodos de aprendizado de dicionário multivisualização tenham sido abordados, eles podem ser melhorados ainda mais. A maioria dos métodos existentes de aprendizado de dicionário multivisualização adota o l0 or l1-restrição de esparsidade da norma nos coeficientes de representação, o que torna as fases de treinamento e teste demoradas. Neste artigo, propomos uma nova abordagem de aprendizado de dicionário multivisualização denominada aprendizagem de dicionários de síntese e análise multivisualização (MSADL), que aprende conjuntamente vários pares de dicionários discriminantes, cada um correspondendo a uma visualização e contendo um dicionário de síntese estruturado e uma análise estruturada. dicionário. MSADL utiliza dicionários de síntese para obter reconstrução específica de classe e usa dicionários de análise para gerar coeficientes de código discriminativos por projeção linear. Além disso, projetamos um termo de não correlação para aprendizagem de dicionário multivisualização, de modo que a redundância entre dicionários de síntese aprendidos a partir de diferentes visualizações possa ser reduzida. Dois conjuntos de dados amplamente utilizados são empregados como dados de teste. Resultados experimentais demonstram a eficiência e eficácia da abordagem proposta.
Fei WU
Nanjing University of Posts and Telecommunications (NJUPT)
Xiwei DONG
Nanjing University of Posts and Telecommunications (NJUPT)
Lu HAN
Nanjing University of Posts and Telecommunications (NJUPT)
Xiao-Yuan JING
Nanjing University of Posts and Telecommunications (NJUPT)
Yi-mu JI
NJUPT
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.
Copiar
Fei WU, Xiwei DONG, Lu HAN, Xiao-Yuan JING, Yi-mu JI, "Multi-View Synthesis and Analysis Dictionaries Learning for Classification" in IEICE TRANSACTIONS on Information,
vol. E102-D, no. 3, pp. 659-662, March 2019, doi: 10.1587/transinf.2018EDL8107.
Abstract: Recently, multi-view dictionary learning technique has attracted lots of research interest. Although several multi-view dictionary learning methods have been addressed, they can be further improved. Most of existing multi-view dictionary learning methods adopt the l0 or l1-norm sparsity constraint on the representation coefficients, which makes the training and testing phases time-consuming. In this paper, we propose a novel multi-view dictionary learning approach named multi-view synthesis and analysis dictionaries learning (MSADL), which jointly learns multiple discriminant dictionary pairs with each corresponding to one view and containing a structured synthesis dictionary and a structured analysis dictionary. MSADL utilizes synthesis dictionaries to achieve class-specific reconstruction and uses analysis dictionaries to generate discriminative code coefficients by linear projection. Furthermore, we design an uncorrelation term for multi-view dictionary learning, such that the redundancy among synthesis dictionaries learned from different views can be reduced. Two widely used datasets are employed as test data. Experimental results demonstrate the efficiency and effectiveness of the proposed approach.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2018EDL8107/_p
Copiar
@ARTICLE{e102-d_3_659,
author={Fei WU, Xiwei DONG, Lu HAN, Xiao-Yuan JING, Yi-mu JI, },
journal={IEICE TRANSACTIONS on Information},
title={Multi-View Synthesis and Analysis Dictionaries Learning for Classification},
year={2019},
volume={E102-D},
number={3},
pages={659-662},
abstract={Recently, multi-view dictionary learning technique has attracted lots of research interest. Although several multi-view dictionary learning methods have been addressed, they can be further improved. Most of existing multi-view dictionary learning methods adopt the l0 or l1-norm sparsity constraint on the representation coefficients, which makes the training and testing phases time-consuming. In this paper, we propose a novel multi-view dictionary learning approach named multi-view synthesis and analysis dictionaries learning (MSADL), which jointly learns multiple discriminant dictionary pairs with each corresponding to one view and containing a structured synthesis dictionary and a structured analysis dictionary. MSADL utilizes synthesis dictionaries to achieve class-specific reconstruction and uses analysis dictionaries to generate discriminative code coefficients by linear projection. Furthermore, we design an uncorrelation term for multi-view dictionary learning, such that the redundancy among synthesis dictionaries learned from different views can be reduced. Two widely used datasets are employed as test data. Experimental results demonstrate the efficiency and effectiveness of the proposed approach.},
keywords={},
doi={10.1587/transinf.2018EDL8107},
ISSN={1745-1361},
month={March},}
Copiar
TY - JOUR
TI - Multi-View Synthesis and Analysis Dictionaries Learning for Classification
T2 - IEICE TRANSACTIONS on Information
SP - 659
EP - 662
AU - Fei WU
AU - Xiwei DONG
AU - Lu HAN
AU - Xiao-Yuan JING
AU - Yi-mu JI
PY - 2019
DO - 10.1587/transinf.2018EDL8107
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E102-D
IS - 3
JA - IEICE TRANSACTIONS on Information
Y1 - March 2019
AB - Recently, multi-view dictionary learning technique has attracted lots of research interest. Although several multi-view dictionary learning methods have been addressed, they can be further improved. Most of existing multi-view dictionary learning methods adopt the l0 or l1-norm sparsity constraint on the representation coefficients, which makes the training and testing phases time-consuming. In this paper, we propose a novel multi-view dictionary learning approach named multi-view synthesis and analysis dictionaries learning (MSADL), which jointly learns multiple discriminant dictionary pairs with each corresponding to one view and containing a structured synthesis dictionary and a structured analysis dictionary. MSADL utilizes synthesis dictionaries to achieve class-specific reconstruction and uses analysis dictionaries to generate discriminative code coefficients by linear projection. Furthermore, we design an uncorrelation term for multi-view dictionary learning, such that the redundancy among synthesis dictionaries learned from different views can be reduced. Two widely used datasets are employed as test data. Experimental results demonstrate the efficiency and effectiveness of the proposed approach.
ER -