Faster nonconvex low-rank matrix learning for image low-level and high-level vision: a unified framework

This study introduces a unified approach to tackle challenges in both low-level and high-level vision tasks for image processing. The framework integrates faster nonconvex low-rank matrix computations and continuity techniques to yield efficient and high-quality results. In addressing real-world ima...

Full description

Saved in:
Bibliographic Details
Main Authors: Zhang, Hengmin, Yang, Jian, Qian, Jianjun, Gong, Chen, Ning, Xin, Zha, Zhiyuan, Wen, Bihan
Other Authors: School of Electrical and Electronic Engineering
Format: Article
Language:English
Published: 2024
Subjects:
Online Access:https://hdl.handle.net/10356/179153
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-179153
record_format dspace
spelling sg-ntu-dr.10356-1791532024-07-22T04:23:32Z Faster nonconvex low-rank matrix learning for image low-level and high-level vision: a unified framework Zhang, Hengmin Yang, Jian Qian, Jianjun Gong, Chen Ning, Xin Zha, Zhiyuan Wen, Bihan School of Electrical and Electronic Engineering Engineering Faster low-rank matrix learning Nonconvex PBCD algorithm This study introduces a unified approach to tackle challenges in both low-level and high-level vision tasks for image processing. The framework integrates faster nonconvex low-rank matrix computations and continuity techniques to yield efficient and high-quality results. In addressing real-world image complexities like noise, variations, and missing data, the framework exploits the intrinsic low-rank structure of the data and incorporates specific residual measurements. The optimization problem for low-rank matrix learning is effectively solved using the nonconvex Proximal Block Coordinate Descent (PBCD) algorithm, resulting in nearly unbiased estimators. Rigorous theoretical analysis ensures both local and global convergence. The PBCD algorithm updates blocks of variables iteratively with closed-form solutions, adeptly handling nonconvexity and promoting faster convergence. Notably, the framework incorporates the randomized singular value decomposition (RSVD) technique and introduces a continuous strategy for adaptive model parameter updates. These strategic choices reduce computational complexity while maintaining result quality. They offer fine-tuned control over the desired rank of the learned matrix and enhance robustness in a straightforward manner. Furthermore, the versatility of the proposed nonconvex PBCD algorithm extends to handling problems with multiple variables, as supported by theoretical analysis. Experimental evaluations, spanning various image low-level and high-level vision tasks such as inpainting, classification, and clustering, validate the effectiveness and efficiency of our framework across diverse databases. The source code is available at https://github.com/ZhangHengMin/FNPBCD_LR. In a nutshell, our framework provides a unified solution to tackle both low-level and high-level vision tasks in images. By combining fast nonconvex low-rank matrix learning with adaptive parameter updates, we achieve efficient computation, yielding high-quality results that demonstrate robustness against various types of noise. The evaluations further endorse the reliability and applicability of our proposed framework. Ministry of Education (MOE) This work was supported in part by the Ministry of Education, Republic of Singapore, through its Start-Up Grant and Academic Research Fund Tier 1 under Grant RG61/22; in part by the National Natural Science Fund (NSF) of China under Grant 61906067, Grant 62176124, Grant 61973162, and Grant 12371510; in part by the China Postdoctoral Science Foundation under Grant 2019M651415 and Grant 2020T130191; in part by the Fundamental Research Funds for the Central Universities under Grant 30918014108, Grant 30920032202, and Grant 30921013114; in part by the NSF of Jiangsu Province under Grant BZ2021013; in part by the NSF for Distinguished Young Scholar of Jiangsu Province under Grant BK20220080; and in part by the ‘‘111’’ Program under Grant B13022. 2024-07-22T04:23:32Z 2024-07-22T04:23:32Z 2024 Journal Article Zhang, H., Yang, J., Qian, J., Gong, C., Ning, X., Zha, Z. & Wen, B. (2024). Faster nonconvex low-rank matrix learning for image low-level and high-level vision: a unified framework. Information Fusion, 108, 102347-. https://dx.doi.org/10.1016/j.inffus.2024.102347 1566-2535 https://hdl.handle.net/10356/179153 10.1016/j.inffus.2024.102347 2-s2.0-85189019031 108 102347 en RG61/22 Information Fusion © 2024 Elsevier B.V. All rights reserved.
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic Engineering
Faster low-rank matrix learning
Nonconvex PBCD algorithm
spellingShingle Engineering
Faster low-rank matrix learning
Nonconvex PBCD algorithm
Zhang, Hengmin
Yang, Jian
Qian, Jianjun
Gong, Chen
Ning, Xin
Zha, Zhiyuan
Wen, Bihan
Faster nonconvex low-rank matrix learning for image low-level and high-level vision: a unified framework
description This study introduces a unified approach to tackle challenges in both low-level and high-level vision tasks for image processing. The framework integrates faster nonconvex low-rank matrix computations and continuity techniques to yield efficient and high-quality results. In addressing real-world image complexities like noise, variations, and missing data, the framework exploits the intrinsic low-rank structure of the data and incorporates specific residual measurements. The optimization problem for low-rank matrix learning is effectively solved using the nonconvex Proximal Block Coordinate Descent (PBCD) algorithm, resulting in nearly unbiased estimators. Rigorous theoretical analysis ensures both local and global convergence. The PBCD algorithm updates blocks of variables iteratively with closed-form solutions, adeptly handling nonconvexity and promoting faster convergence. Notably, the framework incorporates the randomized singular value decomposition (RSVD) technique and introduces a continuous strategy for adaptive model parameter updates. These strategic choices reduce computational complexity while maintaining result quality. They offer fine-tuned control over the desired rank of the learned matrix and enhance robustness in a straightforward manner. Furthermore, the versatility of the proposed nonconvex PBCD algorithm extends to handling problems with multiple variables, as supported by theoretical analysis. Experimental evaluations, spanning various image low-level and high-level vision tasks such as inpainting, classification, and clustering, validate the effectiveness and efficiency of our framework across diverse databases. The source code is available at https://github.com/ZhangHengMin/FNPBCD_LR. In a nutshell, our framework provides a unified solution to tackle both low-level and high-level vision tasks in images. By combining fast nonconvex low-rank matrix learning with adaptive parameter updates, we achieve efficient computation, yielding high-quality results that demonstrate robustness against various types of noise. The evaluations further endorse the reliability and applicability of our proposed framework.
author2 School of Electrical and Electronic Engineering
author_facet School of Electrical and Electronic Engineering
Zhang, Hengmin
Yang, Jian
Qian, Jianjun
Gong, Chen
Ning, Xin
Zha, Zhiyuan
Wen, Bihan
format Article
author Zhang, Hengmin
Yang, Jian
Qian, Jianjun
Gong, Chen
Ning, Xin
Zha, Zhiyuan
Wen, Bihan
author_sort Zhang, Hengmin
title Faster nonconvex low-rank matrix learning for image low-level and high-level vision: a unified framework
title_short Faster nonconvex low-rank matrix learning for image low-level and high-level vision: a unified framework
title_full Faster nonconvex low-rank matrix learning for image low-level and high-level vision: a unified framework
title_fullStr Faster nonconvex low-rank matrix learning for image low-level and high-level vision: a unified framework
title_full_unstemmed Faster nonconvex low-rank matrix learning for image low-level and high-level vision: a unified framework
title_sort faster nonconvex low-rank matrix learning for image low-level and high-level vision: a unified framework
publishDate 2024
url https://hdl.handle.net/10356/179153
_version_ 1806059834458505216