Article
Multidisciplinary Sciences
Sahar Salimpour, Hashem Kalbkhani, Saeed Seyyedi, Vahid Solouk
Summary: In recent years, there has been a growing interest in processing motor imagery electroencephalography (EEG) signals to develop brain-computer interface (BCI) applications. This study presents a semi-supervised model that improves the classification accuracy of motor imagery EEG signals through feature extraction and machine learning algorithms.
SCIENTIFIC REPORTS
(2022)
Review
Computer Science, Artificial Intelligence
Luis Guarda, Juan E. Tapia, Enrique Lopez Droguett, Marcelo Ramos
Summary: This paper presents a Deep Learning-based method for drowsiness detection using CapsNet and spectrogram images of EEG signals. The proposed CapsNet model outperforms the traditional Convolutional Neural Network in terms of accuracy and sensitivity. This method is effective for handling small amounts of data and biomedical signals.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Multidisciplinary Sciences
Menaa Nawaz, Jameel Ahmed
Summary: Real-time data collection and pre-processing enable the automatic assessment of cardiovascular activity through anomaly detection and classification of raw one-dimensional electrocardiogram (ECG) signals. The proposed intelligent end-to-end system utilizes deep learning and multi-label classification algorithms, along with improved feature-engineered parameters and wavelet time scattering features for improved accuracy and robustness. The results show high classification accuracy and F1 score using the long short-term memory (LSTM) method for classification and deep LSTM auto-encoders for anomaly detection.
Article
Engineering, Biomedical
Jiayang Zhang, Kang Li
Summary: In this study, a densely connected convolutional network with multi-view inputs is proposed to decode EEG signals in the Motor Imagery BCI for stroke rehabilitation strategies. The method captures temporal and spatial features based on the whole frequency band and the filtered sub-band signals, and uses dense blocks to enhance model learning capabilities. The proposed method achieves higher accuracy than other deep learning methods.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2023)
Article
Engineering, Civil
Alper Emlek, Murat Peker
Summary: This paper introduces P3SNet, which can generate real-time and competitive disparity maps by using deep convolutional neural networks for end-to-end training from stereo images. The P3SNet architecture consists of two main modules: parallel pyramid pooling and hierarchical disparity aggregation. The parallel pyramid pooling structure allows for intensive extraction of local and global information from multi-scale features, while the hierarchical disparity aggregation provides multi-scale disparity maps using a coarse-to-fine training strategy with the help of costs obtained from multi-scale features.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Fatima Hassan, Syed Fawad Hussain, Saeed Mian Qaisar
Summary: Schizophrenia, a severe mental disorder characterized by disorganized speech and delusions, can be identified using non-invasive and high temporal resolution EEG signals. In this study, a publicly available multi-channel EEG signals dataset is utilized to automatically identify Schizophrenia using a subset of data from selected channels. The combination of three specific channels achieved high accuracies of 90% and 98% on subject-based and non-subject based testing, respectively, using a fusion of CNN and LR.
INFORMATION FUSION
(2023)
Article
Chemistry, Analytical
Zhuo Chen, Dazhi Gao, Kai Sun, Xiaojing Zhao, Yueqi Yu, Zhennan Wang
Summary: In indoor environments, reverberation poses a challenge to sound classification. To overcome this, we used a DenseNet to combine speech spectral features and achieved a classification accuracy of 95.90%, better than other CNNs. The optimized DenseNet model size is only 3.09 MB and can be deployed on embedded devices.
Article
Computer Science, Artificial Intelligence
Domenico Buongiorno, Giacomo Donato Cascarano, Irio De Feudis, Antonio Brunetti, Leonarda Carnimeo, Giovanni Dimauro, Vitoantonio Bevilacqua
Summary: Deep Learning has shown remarkable performance in various tasks such as image recognition, machine translation, and self-driving cars, boosting advancements in physiological signal processing. There has been an exponential increase in studies using DL methods for EMG signal processing, with a focus on hand gesture classification, speech and emotion classification, sleep stage classification, and other applications. The prevalence of convolutional neural networks (CNN) as the most used topology in DL architectures highlights the progress and potential of this field.
Article
Engineering, Biomedical
Ranqi Zhao, Yi Xia, Qiuyang Wang
Summary: This paper proposed a dual-modal and multi-scale deep neural network system using EEG and ECG signals for sleep staging. Experimental results showed that the network exhibited high accuracy in sleep staging.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2021)
Article
Computer Science, Artificial Intelligence
Xue-Yang Min, Kun Qian, Ben-Wen Zhang, Guojie Song, Fan Min
Summary: This study proposes a multi-label active learning algorithm based on serial-parallel neural networks. It addresses the issues of label correlations, missing labels, and label queries through simple and effective mechanisms, achieving state-of-the-art active learning performance.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Biology
Zahra Khademi, Farideh Ebrahimi, Hussain Montazery Kordy
Summary: This study proposes three hybrid models for classifying motor imagery in a brain-computer interface. By using pre-trained convolutional neural networks and long short-term memory networks, combined with transfer learning and data augmentation methods, good classification accuracy is achieved. Among them, the model using Inception-v3 performs the best.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Computer Science, Artificial Intelligence
Lokesh Subramanyan, Udhayakumar Ganesan
Summary: Atrial fibrillation is a common dysrhythmia in healthcare practice with serious consequences. Traditional diagnosis methods have limitations, leading to the need for novel detection models to improve accuracy and efficiency in diagnosis.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Information Systems
Wen-Jie Chen, Jiao Wang, Jian-Qing Li
Summary: Demodulation techniques are crucial for the overall performance of a communication system. This paper proposes a unified architecture for automatic demodulation of modulated signals, using residual units and fully convolutional network. Simulation results demonstrate that the proposed method outperforms existing methods in demodulation performance.
Article
Mathematics
Njud S. Alharbi, Hadi Jahanshahi, Qijia Yao, Stelios Bekiros, Irene Moroz
Summary: This study introduces an ensemble model combining LSTM and CNN models for the classification of ECG signals. The model utilizes LSTM's sequential data learning capability and CNN's intricate pattern recognition strength, along with advanced signal processing methods. Experimental results demonstrate that the proposed model outperforms other deep learning models, showcasing its potential in cardiovascular disease diagnosis.
Article
Engineering, Biomedical
Xuefei Zhao, Dong Liu, Li Ma, Quan Liu, Kun Chen, Shane Xie, Qingsong Ai
Summary: The study introduces a convolutional neural network with an end-to-end serial-parallel structure for analyzing MI-EEG, and improves cross-subject classification accuracy through transfer learning. Experimental results demonstrate that the model performs well in analyzing multi-class MI activities and successfully reduces the number of training parameters.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2022)
Article
Computer Science, Cybernetics
Terence H. Chan, Qi Chen, Raymond W. Yeung
Article
Multidisciplinary Sciences
Wen Qi Zhang, Terence H. Chan, V. Shahraam Afshar
Summary: A correlation propagation model for second-order soliton pulses in the nonlinear Fourier domain is reported for the first time, predicting covariance matrices of soliton pulses at any propagation distance without the need of actually propagating the pulses.
SCIENTIFIC REPORTS
(2021)
Article
Optics
Wen Qi Zhang, Terence H. Chan, Shahraam Afshar
Summary: Nonlinear Fourier transform (NFT) has the potential to overcome capacity limits in fiber optic communication systems, but faces speed and accuracy issues. Machine learning using convolutional neural networks shows promise in NFT-based applications, potentially replacing traditional calculations for decoding information.
Article
Engineering, Electrical & Electronic
Salwa Mostafa, Chi Wan Sung, Guangping Xu, Terence H. Chan
Summary: This paper explores cache placement for ultra-dense fog radio access networks, proposing optimization of cache placement by concatenating MDS codes with repetition codes. By repeating the same packet in some F-APs, multicasting over the fronthaul link can be done in cache placement, saving energy and bandwidth.
IEEE TRANSACTIONS ON COMMUNICATIONS
(2021)
Article
Biochemical Research Methods
Wen Qi Zhang, Alexandra Sorvina, Janna L. Morrison, Jack R. T. Darby, Doug A. Brooks, Sally E. Plush, Shahraam Afshar Vahid
Summary: This study developed a fiber-optic system and two mathematical models for real-time measurement of redox ratios in cells and tissues, which were directly correlated with endogenous fluorescence signals. The results demonstrated the potential application of this system in defining different metabolic disease states.
JOURNAL OF BIOPHOTONICS
(2022)
Article
Engineering, Electrical & Electronic
Ahsan Waqas, Gottfried Lechner, Terence Chan, Khoa Nguyen
Summary: This paper investigates a communication system with digital burst-mode transmission and distributed reception in the presence of carrier frequency offset and Additive White Gaussian Noise (AWGN). The collaboration between nodes is explored to improve the frequency estimation accuracy of low-SNR nodes, and the criteria for selecting and fetching additional symbols are studied. Numerical comparisons and analysis are conducted to evaluate the performance of different schemes.
IET COMMUNICATIONS
(2022)
Article
Engineering, Aerospace
Ahsan Waqas, Khoa Nguyen, Gottfried Lechner, Terence Chan
Summary: This paper presents an algorithm for iterative joint channel parameter estimation and decoding of transmission over channels affected by Doppler shift and Doppler rate using a distributed receiver. The algorithm is derived by applying the sum-product algorithm (SPA) to a factor graph and two methods for dealing with intractable messages of the SPA are proposed.
INTERNATIONAL JOURNAL OF SATELLITE COMMUNICATIONS AND NETWORKING
(2022)
Article
Engineering, Electrical & Electronic
Salwa Mostafa, Chi Wan Sung, Terence H. H. Chan, Guangping Xu
Summary: This study focuses on cooperative caching and delivery in Fog Radio Access Networks (F-RAN), and designs index coding algorithms to minimize fronthaul traffic and transmit energy. The study also considers the tradeoff between fronthaul link traffic load and transmit energy consumption, and crafts algorithms to achieve this tradeoff.
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY
(2022)
Article
Optics
Fengqiu Adam Dong, Wen Qi Zhang, Shaghik Atakaramians, Shahraam Afshar
Summary: Nonreciprocity in optical fibers opens up new possibilities for quantum computing and quantum photonics. This study explores the chiral properties of radiation modes in optical fibers and discovers specific transverse spin angular momenta associated with whispering gallery mode resonances. Through spin-momentum locking, nonreciprocity in the emission coupling of atomic transitions into forward and backward propagating modes is observed and optimized. The findings demonstrate the rich physics and potential applications of fiber radiation modes in light-matter interactions.
OPTICS AND LASER TECHNOLOGY
(2023)
Proceedings Paper
Computer Science, Cybernetics
Ahsan Waqas, Gottfried Lechner, Khoa Nguyen, Terence Chan
Summary: This paper proposes a novel method for joint estimation of multiple parameters using particle filter and fine tuning of particles for faster convergence. Monte Carlo simulations confirm the validity of the proposed algorithm, showing that its performance is close to the ideal scenario in numerical terms.
2021 IEEE LATIN-AMERICAN CONFERENCE ON COMMUNICATIONS (LATINCOM 2021)
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Terence H. Chan, Wenqi Zhang, Sander Wahls, Alan Pak Tao Lau, Shahraam Afshar
2020 IEEE GLOBAL COMMUNICATIONS CONFERENCE (GLOBECOM)
(2020)
Proceedings Paper
Optics
Shahraam Afshar Vahid, Wen Qi Zhang, Terence H. Chan
AOS AUSTRALIAN CONFERENCE ON OPTICAL FIBRE TECHNOLOGY (ACOFT) AND AUSTRALIAN CONFERENCE ON OPTICS, LASERS, AND SPECTROSCOPY (ACOLS) 2019
(2019)
