Article
Computer Science, Hardware & Architecture
Xingyu Yan, Xiaofan Xiong, Xiufeng Cheng, Yujing Huang, Haitao Zhu, Fang Hu
Summary: The HMM-BiMM algorithm combines the Hidden Markov Model and Bi-directional Maximal Matching algorithm to achieve fast and accurate Chinese word segmentation. By dynamically updating the dictionary, it further improves the accuracy and efficiency of word segmentation.
COMPUTERS & ELECTRICAL ENGINEERING
(2021)
Article
Mathematics, Interdisciplinary Applications
Abdessatar Souissi, El Gheteb Soueidi
Summary: This paper aims to expand on previous research on quantum hidden Markov processes by introducing the concept of entangled hidden Markov processes. These are hidden Markov processes in which the hidden processes themselves are entangled Markov processes. The paper provides an explicit expression for the joint expectation of these processes and demonstrates that the approach also applies to the classical case.
CHAOS SOLITONS & FRACTALS
(2023)
Article
Biology
Rui Huo, Liting Zhang, Feifei Liu, Ying Wang, Yesong Liang, Shoushui Wei
Summary: The study proposed a bidirectional hidden semi-Markov model (BI-HSMM) based on the probability distributions of ECG waveform duration for accurate segmentation of ECG waves, achieving excellent performance on the QT database and wearable dynamic electrocardiography (DCG) signals collected by the Shandong Provincial Hospital (SPH). The experimental results and real DCG signal validation confirmed the significant ability of the proposed BI-HSMM method to segment the resting and DCG signals, which is beneficial for detecting and monitoring CVDs.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Computer Science, Artificial Intelligence
Carlos Puerto-Santana, Pedro Larranaga, Concha Bielza
Summary: This article introduces asymmetric hidden Markov models with feature saliencies, which are capable of simultaneously determining relevant variables/features and probabilistic relationships between variables during their learning phase. Comparing with other approaches, the proposed models have better or equal fitness and provide further data insights.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Carlos Puerto-Santana, Pedro Larranaga, Concha Bielza
Summary: In a real-life process evolving over time, the relationship between relevant variables may change. Asymmetric hidden Markov models provide a dynamic framework where different inference models can be used for each state of the process. This paper modifies recent asymmetric hidden Markov models to incorporate an asymmetric autoregressive component for continuous variables, allowing the model to choose the optimal order of autoregression. The paper also demonstrates the adaptation of inference, hidden states decoding, and parameter learning for the proposed model. Experimental results with synthetic and real data showcase the capabilities of this new model.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2022)
Article
Physics, Multidisciplinary
Ozgur Danisman, Umay Uzunoglu Kocer
Summary: Hidden Markov models are commonly used for modeling probabilistic structures with latent variables. They assume that observation symbols are conditionally independent and identically distributed, but in practice, this assumption may not always hold. The proposed model introduces a first-order Markov dependency between the current pair of hidden state-emitted observation symbol and the previous pair, which can better capture possible dependencies in real-life scenarios.
PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS
(2021)
Article
Biochemical Research Methods
Srilakshmi Pattabiraman, Tandy Warnow
Summary: Profile Hidden Markov Models (HMMs) are graphical models that generate finite length sequences from a distribution, widely used in bioinformatics. The construction of profile HMMs is a statistical estimation problem, and it is unknown whether they are statistically identifiable.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2021)
Article
Automation & Control Systems
Rahul Singh, Qinsheng Zhang, Yongxin Chen
Summary: In this paper, an algorithm is proposed for estimating the parameters of a time-homogeneous hidden Markov model (HMM) from aggregate observations. The algorithm is built upon the expectation-maximization algorithm and the aggregate inference algorithm, and it exhibits convergence guarantees for both discrete and continuous observations. When the population size is 1, the algorithm is equivalent to the standard Baum-Welch learning algorithm.
Article
Computer Science, Artificial Intelligence
Isidoros Perikos, Spyridon Kardakis, Ioannis Hatzilygeroudis
Summary: The article introduces a novel, interpretable HMM-based method for recognizing sentiments in text, which has been tested under various architectures, training methods, orders, and ensembles, showing competitive performance and outperforming traditional HMM methods.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Management
Maria Luz Gamiz, Nikolaos Limnios, Maria Del Carmen Segovia-Garcia
Summary: This paper discusses the use of hidden Markov models in reliability engineering, where the state of the system is not directly observable. The paper investigates the maximum-likelihood estimation of system reliability and the effectiveness of preventive maintenance strategies. Extensive simulation studies are conducted to evaluate the finite sample performance of the methodology.
EUROPEAN JOURNAL OF OPERATIONAL RESEARCH
(2023)
Article
Automation & Control Systems
Xiang Zhang, Jun Song, Peng Cheng, Kaibo Shi, Shuping He
Summary: This paper focuses on the asynchronous controller design for directional 2D Markov jump systems (MJSs) based on the two-dimensional Roesser model. The proposed method considers the independent modes of the 2D MJSs in the horizontal and vertical directions and addresses the asynchronous relationship between the system modes and the controller modes. The paper provides a sufficient condition for the mean square exponential stability of the controlled directional 2D MJSs based on Lyapunov theory, while also considering partially unknown transition probabilities in systems and controllers. An example is given to demonstrate the effectiveness of the proposed method.
INTERNATIONAL JOURNAL OF SYSTEMS SCIENCE
(2023)
Article
Chemistry, Multidisciplinary
Anna Borucka, Edward Kozlowski, Rafal Parczewski, Katarzyna Antosz, Leszek Gil, Daniel Pieniak
Summary: Logistics processes and their effective planning, management, and implementation are crucial for businesses. This article analyzes the process of supplying raw materials for production tasks, specifically in a waste processing company where proper management and storage are necessary due to the toxicity of the waste to the environment. The article proposes the use of hidden Markov models, a statistical modeling tool, to assess the level of supply, as traditional methods may not always provide reliable information. The models represent a system as a Markov process with hidden states and visible outputs, with the distribution of outputs defined by a polynomial distribution.
APPLIED SCIENCES-BASEL
(2023)
Article
Computer Science, Information Systems
Laura Symul, Susan Holmes
Summary: In this study, a hierarchical approach based on hidden semi-Markov models is proposed to identify reproductive events and quantify uncertainty in multivariate time series with frequent missing data. The method adapts to changes in tracking behavior, captures variable- and state-dependent missingness, and accurately predicts cycle length.
IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS
(2022)
Article
Biochemistry & Molecular Biology
Ling Wang, John W. Watters, Xiangwu Ju, Genzhe Lu, Shixin Liu
Summary: In this study, a single-molecule assay was used to investigate the role of RNA polymerase collisions on DNA. It was found that collisions between converging RNA polymerases play a vital role in transcription termination and regulation.
Article
Computer Science, Artificial Intelligence
Hui Lan, Ziquan Liu, Janet H. Hsiao, Dan Yu, Antoni B. Chan
Summary: This article proposes a novel HMM-based clustering algorithm, which clusters HMMs through their densities and priors, and simultaneously learns posteriors for the novel HMM cluster centers that compactly represent the structure of each cluster. The numbers K and S are automatically determined in two ways.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Engineering, Electrical & Electronic
Pengpeng Hu, Edmond S. L. Ho, Adrian Munteanu
Summary: This article proposes a novel deep learning framework for generating omnidirectional 3-D point clouds of human bodies by registering front- and back-facing partial scans. The method does not require calibration-assisting devices or assumptions on initial alignment or correspondences. The approach builds virtual correspondences for the partial scans and predicts the rigid transformation between them through deep neural networks. Experiments show that the proposed method achieves state-of-the-art performance in both objective and subjective terms.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2023)
Article
Computer Science, Artificial Intelligence
Qianhui Men, Edmond S. L. Ho, Hubert P. H. Shum, Howard Leung
Summary: In this work, a self-supervised framework called FoCoViL is proposed, which associates actions with common view-invariant properties and simultaneously separates dissimilar viewpoints by maximizing mutual information between multi-view sample pairs. An adaptive focalization method based on pairwise similarity is further proposed to enhance contrastive learning for a clearer cluster boundary. FoCoViL performs well on both unsupervised and supervised classifiers, and the proposed contrastive-based focalization generates a more discriminative latent representation.
Article
Computer Science, Artificial Intelligence
Luca Crosato, Hubert P. H. Shum, Edmond S. L. Ho, Chongfeng Wei
Summary: This paper proposes a framework based on Social Value Orientation and Deep Reinforcement Learning (DRL) for decision-making in the presence of pedestrians. The framework trains decision-making policies with different driving styles using state-of-the-art DRL algorithms in a simulated environment. It also introduces a computationally-efficient pedestrian model suitable for DRL training.
IEEE TRANSACTIONS ON INTELLIGENT VEHICLES
(2023)
Article
Computer Science, Software Engineering
Shuang Chen, Amir Atapour-Abarghouei, Edmond S. L. Ho, Hubert P. H. Shum
Summary: We introduce a software that predicts non-cleft facial images for patients with cleft lip, facilitating the understanding and discussion of cleft lip surgeries. To protect privacy, we design a software framework using image inpainting, which doesn't require cleft lip images for training, mitigating the risk of leakage. We implement a novel multi-task architecture that predicts both non-cleft facial images and facial landmarks, resulting in improved performance as evaluated by surgeons. The software is implemented with PyTorch, supporting consumer-level color images and offering fast prediction speed for effective deployment.
Proceedings Paper
Computer Science, Information Systems
Vidya Rohini Konanur Sathish, Wai Lok Woo, Edmond S. L. Ho
Summary: Identifying sleep stages and patterns is crucial for diagnosing and treating sleep disorders. This paper proposes a CNN architecture to improve the classification performance by benchmarking it against traditional machine learning methods on publicly available sleep datasets. Accuracy, sensitivity, specificity, precision, recall, and F-score are reported as baseline for future research in this direction.
ADVANCES IN CYBERSECURITY, CYBERCRIMES, AND SMART EMERGING TECHNOLOGIES
(2023)
Proceedings Paper
Computer Science, Information Systems
Marco Marchetti, Edmond S. L. Ho
Summary: This paper examines the security issues in Deep Learning and conducts experiments to explore ways to enhance the resilience of DL models against adversarial attacks. The results demonstrate improvements and offer new insights that can guide researchers and practitioners in developing more robust DL algorithms.
ADVANCES IN CYBERSECURITY, CYBERCRIMES, AND SMART EMERGING TECHNOLOGIES
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Li Li, Hubert P. H. Shum, Toby P. Breckon
Summary: This study proposes a semi-supervised semantic segmentation method that achieves superior accuracy with fewer annotations by utilizing a smaller architecture and a novel convolution module. The method also reduces computational costs and improves performance through new data sub-sampling and soft pseudo-label techniques.
2023 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR)
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Ziyi Chang, Edmund J. C. Findlay, Haozheng Zhang, Hubert P. H. Shum
Summary: This article proposes a denoising diffusion probabilistic model solution for generating styled motion of digital humans. By representing both inter-class motion content and intra-class style behavior in the same latent, an integrated, end-to-end trained pipeline is achieved. A multi-task architecture of diffusion model and adversarial and physical regulations are designed, resulting in superior performance.
PROCEEDINGS OF THE 18TH INTERNATIONAL JOINT CONFERENCE ON COMPUTER VISION, IMAGING AND COMPUTER GRAPHICS THEORY AND APPLICATIONS, VISIGRAPP 2023
(2023)
Article
Computer Science, Interdisciplinary Applications
Dimitrios Sakkos, Edmond S. L. Ho, Hubert P. H. Shum, Garry Elvin
Summary: The researchers addressed the challenge of illumination changes in background subtraction using data augmentation and proposed a post-processing method to improve the accuracy of segmentation. The experiments demonstrated the significant contribution of this method in handling illumination changes.
JOURNAL OF ENTERPRISE INFORMATION MANAGEMENT
(2023)
Proceedings Paper
Computer Science, Interdisciplinary Applications
Haozheng Zhang, Edmond S. L. Ho, Xiatian Zhang, Hubert P. H. Shum
Summary: Parkinson's disease is a progressive neurodegenerative disorder with challenging diagnosis. We propose a low-cost Parkinson's tremor classification system using video recording of human movements, which incorporates an attention module to extract relevant information and filter noise. Experimental results show superior performance of our system.
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION, MICCAI 2022, PT IV
(2022)