Incoherent dictionary pair learning: application to a novel open-source database of Chinese numbers

Vahid Abolghasemi; Mingyang Chen; Ali Alameer; Saideh Ferdowsi; Jonathon Chambers; Kianoush Nazarpour

doi:10.1109/LSP.2018.2798406

Incoherent dictionary pair learning: application to a novel open-source database of Chinese numbers

Vahid Abolghasemi, Mingyang Chen, Ali Alameer, Saideh Ferdowsi, Jonathon Chambers, Kianoush Nazarpour

Research output: Contribution to journal › Article › peer-review

Abstract

We enhance the efficacy of an existing dictionary pair learning algorithm by adding a dictionary incoherence penalty term. After presenting an alternating minimization solution, we apply the proposed incoherent dictionary pair learning (InDPL) method in classification of a novel open-source database of Chinese numbers. Benchmarking results confirm that the InDPL algorithm offers enhanced classification accuracy, especially when the number of training samples is limited.

Original language	English
Pages (from-to)	472 - 476
Number of pages	5
Journal	IEEE Signal Processing Letters
Volume	25
Issue number	4
Early online date	25 Jan 2018
DOIs	https://doi.org/10.1109/LSP.2018.2798406
Publication status	Published - 1 Apr 2018

Keywords / Materials (for Non-textual outputs)

Chinese numbers
classification
incoherent dictionary pair learning

Access to Document

10.1109/LSP.2018.2798406Licence: All Rights Reserved

Cite this

@article{913f3ecf48e040d78da40479d73c1edc,

title = "Incoherent dictionary pair learning: application to a novel open-source database of Chinese numbers",

abstract = "We enhance the efficacy of an existing dictionary pair learning algorithm by adding a dictionary incoherence penalty term. After presenting an alternating minimization solution, we apply the proposed incoherent dictionary pair learning (InDPL) method in classification of a novel open-source database of Chinese numbers. Benchmarking results confirm that the InDPL algorithm offers enhanced classification accuracy, especially when the number of training samples is limited.",

keywords = "Chinese numbers, classification, incoherent dictionary pair learning",

author = "Vahid Abolghasemi and Mingyang Chen and Ali Alameer and Saideh Ferdowsi and Jonathon Chambers and Kianoush Nazarpour",

year = "2018",

month = apr,

day = "1",

doi = "10.1109/LSP.2018.2798406",

language = "English",

volume = "25",

pages = "472 -- 476",

journal = "IEEE Signal Processing Letters",

issn = "1070-9908",

publisher = "Institute of Electrical and Electronics Engineers",

number = "4",

}

TY - JOUR

T1 - Incoherent dictionary pair learning: application to a novel open-source database of Chinese numbers

AU - Abolghasemi, Vahid

AU - Chen, Mingyang

AU - Alameer, Ali

AU - Ferdowsi, Saideh

AU - Chambers, Jonathon

AU - Nazarpour, Kianoush

PY - 2018/4/1

Y1 - 2018/4/1

N2 - We enhance the efficacy of an existing dictionary pair learning algorithm by adding a dictionary incoherence penalty term. After presenting an alternating minimization solution, we apply the proposed incoherent dictionary pair learning (InDPL) method in classification of a novel open-source database of Chinese numbers. Benchmarking results confirm that the InDPL algorithm offers enhanced classification accuracy, especially when the number of training samples is limited.

AB - We enhance the efficacy of an existing dictionary pair learning algorithm by adding a dictionary incoherence penalty term. After presenting an alternating minimization solution, we apply the proposed incoherent dictionary pair learning (InDPL) method in classification of a novel open-source database of Chinese numbers. Benchmarking results confirm that the InDPL algorithm offers enhanced classification accuracy, especially when the number of training samples is limited.

KW - Chinese numbers

KW - classification

KW - incoherent dictionary pair learning

UR - https://doi.org/10.1109/LSP.2018.2798406

U2 - 10.1109/LSP.2018.2798406

DO - 10.1109/LSP.2018.2798406

M3 - Article

SN - 1070-9908

VL - 25

SP - 472

EP - 476

JO - IEEE Signal Processing Letters

JF - IEEE Signal Processing Letters

IS - 4

ER -

Incoherent dictionary pair learning: application to a novel open-source database of Chinese numbers

Abstract

Keywords / Materials (for Non-textual outputs)

Access to Document

Other files and links

Fingerprint

Cite this