Article
Automation & Control Systems
Haishan Ye, Tong Zhang
Summary: This paper explores the problem of decentralized principal components analysis (PCA) and proposes an algorithm called DeEPCA, which has a convergence rate similar to centralized PCA but with the best communication complexity among existing decentralized PCA algorithms. DeEPCA is the first decentralized PCA algorithm where the number of communication rounds for each power iteration is independent of target precision.
JOURNAL OF MACHINE LEARNING RESEARCH
(2021)
Article
Computer Science, Artificial Intelligence
Yuan Tai, Yihua Tan, Shengzhou Xiong, Jinwen Tian
Summary: The correlation filter achieves high performance in dealing with motion blur or lighting changes, but faces challenges in scenarios with occlusion. This study proposes the subspace reconstruction based CF (SRBCF) tracker to address the issue of background interference, achieving significant improvements in tracking accuracy.
COMPUTER VISION AND IMAGE UNDERSTANDING
(2021)
Article
Computer Science, Hardware & Architecture
Ahmad Saeed Khattak, Nadeem Anjum, Nasrullah Khan, Muhammad Rafiq Mufti, Naeem Ramzan
Summary: Surveillance systems use object tracking methods to locate targets accurately. Object tracking algorithms need to track under various constraints for a long time and allocate system resources for higher-level tasks. This study focuses on detection-free tracking and evaluates the AMF-MSPF algorithm through experiments while highlighting some areas for improvement.
Article
Computer Science, Artificial Intelligence
Chao Ma, Minjie Wan, Yunkai Xu, Kan Ren, Weixian Qian, Qian Chen, Guohua Gu
Summary: This paper proposes an infrared target tracker based on the proximal robust principal component analysis method, solves the convex optimization problem using the Alternating Direction Method of Multipliers, and locates the target using the particle filter framework.
APPLIED INTELLIGENCE
(2022)
Article
Geochemistry & Geophysics
Yangfan Li, Chunjiang Bian, Hongzhen Chen
Summary: This study proposes a satellite video object tracking algorithm that improves robustness through particle sampling and motion estimation, enhances robustness to target rotation through the use of a color histogram model, and effectively handles background clutter and low contrast by fusing multiple feature response maps.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Review
Computer Science, Information Systems
Alex Marchioni, Luciano Prono, Mauro Mangia, Fabio Pareschi, Riccardo Rovatti, Gianluca Setti
Summary: Subspace analysis is widely used for high-dimensional data and is essential for early treatment of signal-processing tasks. Traditional methods require significant resources, but specialized streaming algorithms allow SA to run efficiently on low-power devices. This article provides a classification, comparison, and evaluation of these methods for subspace identification.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Engineering, Electrical & Electronic
Arpita Gang, Bingqing Xiang, Waheed U. Bajwa
Summary: This paper discusses the problem of distributed PSA/PCA in a network of machines without a central server, proposing two algorithms and analyzing their convergence in the case of sample-wise partitioned data. Extensive experiments are conducted to validate the usefulness of the proposed algorithms, including an MPI-based distributed implementation to study network topology, communication cost, and the effects of straggler machines.
IEEE TRANSACTIONS ON SIGNAL AND INFORMATION PROCESSING OVER NETWORKS
(2021)
Article
Computer Science, Artificial Intelligence
Tianyue Zhang, Furao Shen, Tao Zhu, Jian Zhao
Summary: This article proposes a fast and robust incremental subspace learning framework called EOCA for dimensionality reduction. EOCA determines the target dimensionality automatically by adaptive thresholds and extracts the low-dimensional representation of data in the form of orthogonal subspace bases. Additionally, EOCA provides the ability to merge subspaces and eliminate outlier effects.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Yi Zhang, Guixi Liu, Hanlin Huang, Ruke Xiong
Summary: This paper proposes a unified tracking framework that combines lightweight Siamese network and template-guided learning to address the balance between high performance and real-time execution on resource-constrained devices. By using a two-step pruning method to compress the Siamese network and constructing a template-guided learning model, the proposed method achieves efficient tracking performance.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Interdisciplinary Applications
Won-Jae Lee, Myung-Il Roh, Hye-Won Lee, Jisang Ha, Yeong-Min Cho, Sung-Jun Lee, Nam-Sun Son
Summary: In this study, ship awareness was achieved through ship detection using YOLO(v3), constructing a virtual image dataset using Unity, calculating the position of the horizon in ship images, tracking target ships' positions and speeds, and proposing a deep learning model to determine ship headings. The proposed method for ship awareness based on camera images was successfully validated in various scenarios, demonstrating good detection, localization, and tracking performance.
JOURNAL OF COMPUTATIONAL DESIGN AND ENGINEERING
(2021)
Article
Engineering, Electrical & Electronic
Baojie Fan, Yang Cong, Yandong Tang, Jiandong Tian, Chenliang Xu
Summary: Our proposed method assigns different channel features to different kernel spaces to enhance discriminability, divides targets into multi-layer patches for subtask correlation filtering, and incorporates globally and locally structured correlation filters to complement each other with similar motion models.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2021)
Article
Economics
Florian Huber, Gary Koop
Summary: This paper proposes a conjugate Bayesian model that combines vector autoregression (VAR) and factor models by shrinking towards the subspace defined by the factor model. The approach allows for estimating the strength of shrinkage and the number of factors. The paper demonstrates the effectiveness of the method in detecting the number of factors and improving macroeconomic forecast using simulations and US macroeconomic data.
JOURNAL OF APPLIED ECONOMETRICS
(2023)
Article
Biochemical Research Methods
Charles C. David, Chris S. Avery, Donald J. Jacobs
Summary: JEDi software is an upgraded tool that employs multithreading and user-friendly interface for rapid investigation of conformational motions of biopolymers, including multiple chain proteins. It offers options for Cartesian-based coordinates (cPCA) and internal distance pair coordinates (dpPCA) to construct covariance, correlation, and partial correlation matrices.
BMC BIOINFORMATICS
(2021)
Article
Neurosciences
Christian Merkel, Jens-Max Hopf, Mircea Ariel Schoenfeld
Summary: Previous work has shown different markers for location-based or object-based representations of tracked targets. This study aims to identify the spectral properties of the electrophysiological signal during tracking that might indicate these different representations. The results indicate that object-based tracking is supported by enhanced encoding during the initial presentation of the targets, while location-based tracking is characterized by sustained maintenance of the individual targets throughout the tracking period.
EUROPEAN JOURNAL OF NEUROSCIENCE
(2022)
Article
Engineering, Industrial
Arthur Henrique de Andrade Melani, Miguel Angelo de Carvalho Michalski, Renan Favarao da Silva, Gilberto Francisco Martha de Souza
Summary: Through Condition-Based Maintenance strategy, a hybrid framework based on Moving Window Principal Component Analysis (MWPCA) and Bayesian Network (BN) was proposed for automated Fault Detection and Diagnosis (FDD) in machinery. The framework was able to detect and diagnose several simulated failures in a simplified model of a hydrogenerator.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2021)