Dictionary learning for signals in additive noise with generalized Gaussian distribution

被引：2

作者：

Zheng, Xiaomeng ^{[1
]}

Dumitrescu, Bogdan ^{[2
]}

Liu, Jiamou ^{[3
]}

Giurcaneanu, Ciprian Doru ^{[1
]}

机构：

[1] Univ Auckland, Dept Stat, Private Bag 92019, Auckland 1142, New Zealand

[2] Univ Politehn Bucuresti, 313 Spl Independentei, Bucharest 060042, Romania

[3] Univ Auckland, Sch Comp Sci, Auckland, New Zealand

来源：

SIGNAL PROCESSING | 2022年 / 195卷

关键词：

Adaptive dictionary learning; Generalized Gaussian distribution; Information theoretic criteria; Shape parameter; Image data; ALGORITHM; DESIGN; SVD;

D O I：

10.1016/j.sigpro.2022.108488

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

We propose a dictionary learning (DL) algorithm for signals in additive noise with generalized Gaussian distribution (GGD) by redesigning three key components used in DL for Gaussian signals: (i) the orthog-onal matching pursuit algorithm, (ii) the approximate K-SVD algorithm and (iii) the information theoretic criteria. In experiments with simulated data, we show that the performance of the new algorithm is higher or equal to the performance of the DL algorithms for signals in Laplacian noise. We also discuss how the shape parameter of the GGD noise can be estimated. For image data, we examine the relation-ship between the complexity of the DL model and the errors obtained on the test set. This provides guidance on the values of the shape parameter that should be employed in image modeling.(c) 2022 Elsevier B.V. All rights reserved.

引用

页数：12

共 50 条

[21] OPTIMUM DETECTION OF NON-GAUSSIAN SIGNALS AGAINST THE BACKGROUND OF ADDITIVE NOISE
BROVKOV, BA
SHLYAKHIN, VM
RADIOTEKHNIKA I ELEKTRONIKA, 1984, 29 (07): : 1420 - 1421
[22] Capacity of Peak and Average Power Constrained Additive Generalized Gaussian Noise Channels
Narayanan, Prashant
Theagarajan, Lakshmi Narasimhan
IEEE COMMUNICATIONS LETTERS, 2022, 26 (12) : 2880 - 2883
[23] Non-Orthogonal Multiple Access in the Presence of Additive Generalized Gaussian Noise
Bariah, Lina
Muhaidat, Sami
Sofotasios, Paschalis C.
Gurugopinath, Sanjeev
Hamouda, Walaa
Yanikomeroglu, Halim
IEEE COMMUNICATIONS LETTERS, 2020, 24 (10) : 2137 - 2141
[24] Identification of the distribution parameters of additive and multiplicative non-Gaussian noise
Artyushenko V.M.
Volovach V.I.
Artyushenko, V.M. (artuschenko@mail.ru), 2017, Allerton Press Incorporation (53) : 230 - 236
[25] Energy Detection of Gaussian Signals Subject to Impulsive Noise in Generalized Fading Channels
de Miranda Cardoso, Jose Vinicius
Lira Queiroz, Wamberto Jose
Liu, Hang
de Alencar, Marcelo Sampaio
WIRELESS ALGORITHMS, SYSTEMS, AND APPLICATIONS, WASA 2016, 2016, 9798 : 231 - 240
[26] Synthesis of detectors of signals backgrounded by noise with the generalized exponential distribution
Gordienko, V. I.
Radioelectronics and Communications Systems, 38 (06):
[27] Structured Dictionary Learning for Image Denoising Under Mixed Gaussian and Impulse Noise
Zhu, Hong
Ng, Michael K.
IEEE TRANSACTIONS ON IMAGE PROCESSING, 2020, 29 : 6680 - 6693
[28] Robust Bayesian non-parametric dictionary learning with heterogeneous Gaussian noise
Wang, Yi
Li, Bin
Wang, Yang
Chen, Fang
Zhang, Bang
Li, Zhidong
COMPUTER VISION AND IMAGE UNDERSTANDING, 2016, 150 : 31 - 43
[29] Compressive detection of sparse signals in additive white Gaussian noise without signal reconstruction
Hariri, Alireza
Babaie-Zadeh, Massoud
SIGNAL PROCESSING, 2017, 131 : 376 - 385
[30] Bias and variance of the PWVD based IF estimator for polynomial FM signals in additive Gaussian noise
Barkat, B
Boashash, B
NINTH IEEE SIGNAL PROCESSING WORKSHOP ON STATISTICAL SIGNAL AND ARRAY PROCESSING, PROCEEDINGS, 1998, : 356 - 359

← 1 2 3 4 5 →