Abhay Upadhyay - Profile on Academia.edu (original) (raw)

Papers by Abhay Upadhyay

International Journal For Multidisciplinary Research, Jan 5, 2024

Traffic signal control plays a vital role in effectively managing traffic flow and alleviating co... more Traffic signal control plays a vital role in effectively managing traffic flow and alleviating congestion in urban areas. Traditional methods for controlling traffic signals often rely on fixed timing plans or predefined algorithms, which may not be adaptable to changing traffic conditions. Reinforcement Learning is gaining traction as a favored data-centric method for adapting traffic signal control in intricate urban traffic networks. This article represents a conceptual review of recent studies and techniques that showcase the effectiveness of Deep Reinforcement Learning (DRL) in enhancing the performance of traffic signal control. These improvements include reducing travel time, fuel consumption, and emissions. Additionally, we will delve into different algorithms and learning systems explored in research papers, such as multi-agent reinforcement learning and Deep Q Networks (DQN).

Automated Variational Nonlinear Chirp Mode Decomposition for Bearing Fault Diagnosis

IEEE Transactions on Industrial Informatics

Sleep Stage Classification Using DWT and Dispersion Entropy Applied on EEG Signals

Computer-aided Design and Diagnosis Methods for Biomedical Applications, 2021

New methods based on variational mode decomposition for speech signal analysis

Empirical wavelet transform based classification of surface electromyogram signals for hand movements

Automated diagnosis of muscle diseases from EMG signals using empirical mode decomposition based method

Biomedical Signal Processing and Control, 2022

Abstract Muscle activity decreases due to various conditions like age factors and muscle diseases... more Abstract Muscle activity decreases due to various conditions like age factors and muscle diseases namely, amyotrophic lateral sclerosis (ALS) and myopathy. Electromyogram (EMG) signals are regularly explored by specialists to analyze the irregularity of muscles. Manual investigation of EMG signals is a tedious task for medical practitioners. Therefore, this work proposes a new method for classifying the ALS, myopathy, and normal EMG signals. In the proposed method, the empirical mode decomposition (EMD) method is applied to decompose the EMG signals into intrinsic mode functions (IMFs). The suitable IMFs for feature selection are selected using the t-test based approach and used to compute the foot distances denoted as fp 1 and fp 2 by constructing the complex plane plot. The quadrilateral is drawn over a complex plot by considering fp 1 and fp 2 as a diagonal of it, followed by calculating the area (A) and circumference (CF) of the quadrilateral. These measures are utilized for separating the three classes of myopathy, ALS, and normal EMG signals. The proposed algorithm has been trained and validated using a feed forward neural network (FFNN), support vector machine (SVM), and decision tree. The algorithm, when tested with a FFNN, achieved the maximum classification accuracy, sensitivity, and specificity of 99.53%, 99.25% and 99.60%, respectively.

Circuits, Systems, and Signal Processing, 2020

In this paper, a novel method is presented to analyze the amplitude modulated and frequency modul... more In this paper, a novel method is presented to analyze the amplitude modulated and frequency modulated (AM-FM) multicomponent signals using a combination of the variational mode decomposition (VMD) and the discrete energy separation algorithm (DESA). In the presented method, firstly, a multicomponent signal is decomposed using VMD method applied in an iterative way. In order to separate the monocomponent signals from multicomponent signal, a suitable convergence criterion is developed based on the values of estimated center frequencies (CF) and standard deviations (σ CF ) of the decomposed components. Further, the estimation of amplitude envelope and the instantaneous frequency functions of monocomponent AM-FM signals has been carried out by employing DESA. Moreover, the proposed method is also applied on the synthetic AM-FM signal and speech signals to evaluate its performance. Furthermore, its performance is also compared with the Fourier-Bessel series expansion-based DESA, empirical wavelet transform-based DESA, and iterative eigenvalue decomposition-based DESA methods. The performance of the proposed method is compared with the other methods in terms of mean square error between actual and estimated amplitude envelopes (MSE AE ), mean square error between actual and estimated instantaneous frequencies (MSE IF ) for synthetic signal. The COSH distance measure is used as a performance measure for speech signals. It is found that the proposed method gives better results in terms of performance measures in several cases.

Computers & Electrical Engineering, 2019

The patient monitoring and diagnosis using electrocardiogram (ECG) signal involve accurate QRS co... more The patient monitoring and diagnosis using electrocardiogram (ECG) signal involve accurate QRS complex detection. However, the reliability of QRS detection is limited by the presence of variety of noises and morphologies. A high performance QRS complex detection scheme based on the tunable-Q wavelet transform (TQWT) is presented in this paper. The tunable parameters of TQWT can improve the localization of QRS when selected suitably. These parameters help to choose appropriate mother wavelet and proper sub-band of interest with fine separation from the undesired sub-bands. The method starts with TQWT based decomposition and reconstruction of the ECG signal with optimally selected input parameters and sub-bands. The correntropy envelope is computed from the resultant QRS enhanced signal which in turn assists the QRS localization. Our proposed method has yielded an average detection accuracy of 99.88%, sensitivity of 99.89% and positive productivity of 99.83% on the MIT-BIH arrhythmia database.

Computers & Electrical Engineering, 2017

In this paper, a novel approach is proposed to determine instantaneous fundamental frequency (IFF... more In this paper, a novel approach is proposed to determine instantaneous fundamental frequency (IFF) of speech signals. In the proposed method, the detected voiced speech signals are filtered into low frequency range (LFR) and divided into smaller segments. Further, the fundamental frequency components (FFCs) of each segment of the LFR filtered voiced signals have been obtained using the variational mode decomposition method used in iterative way with suitable convergence criteria. In order to determine IFF, the Hilbert transform and smoothing operation have been applied on the obtained FFC. The FFC is obtained by concatenating all the extracted FFCs corresponding to each LFR filtered voiced speech segment. The performance of the proposed method has been evaluated in different databases and also compared with some other existing methods in the presence of white, babble, and vehicular noises with different signal to noise ratios.

Future Generation Computer Systems, 2019

To perform basic hand movements, a hand amputee person needs an exoskeleton prosthetic hand (EPH)... more To perform basic hand movements, a hand amputee person needs an exoskeleton prosthetic hand (EPH). The EPH can be controlled through electroencephalogram (EEG) or electromyogram (EMG) signals. The EMG signals are preferred as they are acquired from surface of forearm and termed as surface EMG (sEMG). It is very challenging to design the control section for EPH. It should be able to classify different hand movements accurately based on the acquired sEMG signals. Also the sEMG signals must be acquired from minimum number of electrodes to make EPH cost-effective. In this paper, we have proposed a novel technique to classify the basic hand move-

Electronics Letters, 2017

A novel and effective speech enhancement method is proposed for suppressing the white noise as we... more A novel and effective speech enhancement method is proposed for suppressing the white noise as well as non-stationary acoustic noises. The proposed method employs the combination of variational mode decomposition (VMD) and empirical mode decomposition (EMD) methods. In this method, first EMD is used to decompose the noisy speech signal into intrinsic mode functions (IMFs). Further, the VMD is applied on summation of selected IMFs. The main contribution of the proposed method is the selection of IMFs based on Hurst exponent and further applying steps of VMD method. The proposed modified EMD-VMD (mEMD-VMD) method is suitable to reduce the low-frequency noise as well as high-frequency noise. The proposed method gives the better results in terms of speech quality and composite measures. The proposed study is performed on eight speech signals under additive white Gaussian noise, street noise, babble noise, and airport noise taken from NOIZEUS database.

Applied Sciences, 2017

This paper analyzes the underlying complexity and non-linearity of electroencephalogram (EEG) sig... more This paper analyzes the underlying complexity and non-linearity of electroencephalogram (EEG) signals by computing a novel multi-scale entropy measure for the classification of seizure, seizure-free and normal EEG signals. The quality factor (Q) based multi-scale entropy measure is proposed to compute the entropy of the EEG signal in different frequency-bands of interest. The Q -based entropy (QEn) is computed by decomposing the signal with the tunable-Q wavelet transform (TQWT) into the number of sub-bands and estimating K-nearest neighbor (K-NN) entropies from various sub-bands cumulatively. The optimal selection of Q and the redundancy parameter (R) of TQWT showed better robustness for entropy computation in the presence of high-and low-frequency components. The extracted features are fed to the support vector machine (SVM) classifier with the wrapper-based feature selection method. The proposed method has achieved accuracy of 100% in classifying normal (eyes-open and eyes-closed) and seizure EEG signals, 99.5% in classifying seizure-free EEG signals (from the hippocampal formation of the opposite hemisphere of the brain) from seizure EEG signals and 98% in classifying seizure-free EEG signals (from the epileptogenic zone) from seizure EEG signals, respectively, using the SVM classifier. We have also achieved classification accuracies of 99% and 98.6% in classifying seizure versus non-seizure EEG signals and the individual three classes, namely normal, seizure-free and seizure EEG signals, respectively. The performance measure of the proposed multi-scale entropy has been found to be comparable with the existing state of the art epileptic EEG signals classification methods studied using the same database.

Neural Computing and Applications, 2017

Computer-aided sleep monitoring system can effectively reduce the burden of experts in analyzing ... more Computer-aided sleep monitoring system can effectively reduce the burden of experts in analyzing the large volume of electroencephalogram (EEG) recordings corresponding to sleep stages. In this paper, a new technique for automated classification of sleep stages based on iterative filtering of EEG signals is presented. In order to perform sleep stages classification, the EEG signals are decomposed using iterative filtering method. The modes obtained from iterative filtering of EEG signal can be considered as amplitude-modulated and frequency-modulated (AM-FM) components. The discrete energy separation algorithm (DESA) is applied to the modes to determine amplitude envelope and instantaneous frequency functions. The extracted amplitude envelope and instantaneous frequency functions have been used to compute Poincare ´plot descriptors and statistical measures. The Poincare ´plot descriptors and statistical measures are applied as input features for different classifiers in order to classify sleep stages. The classifiers namely, naı ¨ve Bayes, k-nearest neighbor, multilayer perceptron, C4.5 decision tree, and random forest are applied in order to classify the EEG epochs corresponding to various sleep stages. The experimental study has been performed on online available Sleep-EDF database for two-class to six-class classification of sleep stages based on EEG signals. The two-class to six-class classification problems are formulated by taking different combinations of EEG signals corresponding to various sleep stages. The comparison of the results is presented for different multi-class classification problems with the other recently proposed methods. The results show that the proposed method has provided better tenfold crossvalidation classification accuracy than other existing methods.

An integrated alcoholic index using tunable-Q wavelet transform based features extracted from EEG signals for diagnosis of alcoholism

Applied Soft Computing, 2017

Graphical abstractDisplay Omitted HighlightsWe propose a new method for diagnosis of alcoholism u... more Graphical abstractDisplay Omitted HighlightsWe propose a new method for diagnosis of alcoholism using TQWT.New feature set based on correntropy derived from TQWT have been proposed.The effects of Q on classification performance have been evaluated.A novel Alcoholism Risk Index (ARI) is developed using 3 clinically significant features.Performance has been compared with existing methods. Alcoholism affects the structure and functioning of brain. Electroencephalogram (EEG) signals can depict the state of brain. The EEG signals are ensemble of various neuronal activity recorded from different scalp regions having different characteristics and very low magnitude in microvolts. These factors make human interpretation difficult and time consuming to analyze these signals. Moreover, these highly varying EEG signals are susceptible to inter/intra variability errors. So, a Computer-Aided Diagnosis (CAD) can be used to identify the alcoholic and normal subjects accurately. However, these EEG signals exhibit nonlinear and non-stationary properties. Therefore, it needs much effort in deciphering the diagnostic evidence from them using linear time and frequency-domain methods. The nonlinear parameters together with time-frequency/scale domain methods can help to detect tiny changes in these signals. The correntropy is nonlinear indicator which characterizes the dynamic behavior of EEG signals in time-scale domain. In this paper, we present a new way for diagnosis of alcoholism using Tunable-Q Wavelet Transform (TQWT) based features derived from EEG signals. The feature extraction is performed using TQWT based decomposition and extracted Centered Correntropy (CC) from the forth decomposed detail sub-band. The Principal Component Analysis (PCA) is used for feature reduction followed by Least Squares-Support Vector Machine (LS-SVM) for classifying normal and alcoholic EEG signals. In order to make sure reliable classification performance, 10-fold cross-validation scheme is adopted. Our proposed system is able to diagnose the alcoholic and normal EEG signals, with an average accuracy of 97.02%, sensitivity of 96.53%, specificity of 97.50% and Matthews correlation coefficient of 0.9494 for Q-factor (Q) varying between 3 and 8 using Radial Basis Function (RBF) kernel function. Also, we have established a novel Alcoholism Risk Index (ARI) using three clinically significant features to discriminate the given classes by means of a single number. This system can be used for automated diagnosis and monitoring of alcoholic subjects to evaluate the effect of treatment.

Detection of glottal closure instants from voiced speech signals using the Fourier-Bessel series expansion

2015 International Conference on Communications and Signal Processing (ICCSP), 2015

Glottal closure instants (GCIs) present in the voiced speech is the term used to define the insta... more Glottal closure instants (GCIs) present in the voiced speech is the term used to define the instants of significant excitation of the vocal tract system during the production of the speech. In this paper, a novel method for GCI detection via Fourier-Bessel (FB) series expansion of the voiced speech signal in low frequency range (LFR) is explored. GCIs in the proposed method are detected as the local minima of the fundamental frequency (F0) component obtained from the voiced speech signal in LFR. The proposed method has been studied for male and female speech signals from CMU-Arctic database. Experimental results are included in order to show the performance of the proposed method for GCI detection from speech signals.

Journal of the Franklin Institute, 2015

In this paper, a variational mode decomposition (VMD) based method has been proposed for the inst... more In this paper, a variational mode decomposition (VMD) based method has been proposed for the instantaneous detection of voiced/non-voiced (V/NV) regions in the speech signals. In the proposed method, the VMD is applied in iterative way with specific input parameters. Firstly, the VMD decomposes the speech signal into two components, then, the VMD is applied successively on one of these two components based on suitably defined convergence criteria. It has been shown that the VMD applied in iterative way behaves as a low-pass filter and after convergence it provides separation of the fundamental frequency (F 0 ) component from the speech signal. The envelope of the F 0 component of the speech signal has been obtained using an analytical model based on single degree of freedom (SDOF). Automatic threshold has been computed from the obtained envelope in order to detect the V/NV regions in speech signals. The proposed method has been studied on speech signals and the corresponding electroglottograph (EGG) signals from the CMU-Arctic database in different noise conditions obtained from the NOISEX-92 database. Experimental results at various signal to noise ratios (SNRs) are included in order to show the effectiveness of the proposed method compared to the other existing methods for V/NV detection in speech signals.

Integrating shape and Edge Histogram Descriptor with Stationary Wavelet Transform for Effective Content Based Image Retrieval

2014 International Conference on Circuits, Power and Computing Technologies [ICCPCT-2014], 2014

In this paper we propose a hybrid approach for Effective Content Based Image Retrieval based on t... more In this paper we propose a hybrid approach for Effective Content Based Image Retrieval based on texture and shape feature. Towards this, first Stationary Wavelet Transform (SWT) is applied on query image to extract horizontal, vertical and diagonal detail matrices. Stationary Wavelet Transform is used because of its translational invariant property. After this Edge Histogram Descriptor (EHD) is used to exploit the absolute location of edges in the image as well their global composition. To aid the retrieval process, five different shape measures have also been included. Finally Euclidean distance is used to retrieve the relevant results. Experimental results show that the combination of SWT and EHD techniques provides significant improvement over existing methods thereby increasing the retrieval efficiency.

Fusion of local and global features using Stationary Wavelet Transform for efficient Content Based Image Retrieval

2014 IEEE Students' Conference on Electrical, Electronics and Computer Science, 2014

In this paper we propose a hybrid approach for Content Based Image Retrieval that takes into acco... more In this paper we propose a hybrid approach for Content Based Image Retrieval that takes into account both global as well as local features of an image. Towards this, first Stationary Wavelet Transform is applied on query image to extract horizontal, vertical and diagonal detail matrices. Stationary Wavelet Transform is used because of its translational invariant property. After this global textural features are extracted using Gray level Co-occurrence Matrix for each of these sub-matrices. To aid the retrieval process, a local descriptor is also computed by splitting the image into sub-regions. Finally Euclidean distance is used to retrieve the relevant results. Experimental results show that the proposed approach provides significant improvement over existing methods.

A new method for determination of instantaneous pitch frequency from speech signals

2015 IEEE Signal Processing and Signal Processing Education Workshop (SP/SPE), 2015

This paper presents a new method for instantaneous pitch frequency determination from speech sign... more This paper presents a new method for instantaneous pitch frequency determination from speech signals. The proposed method is based on the variational mode decomposition (VMD) and the Hilbert transform. The VMD is applied in iterative way with specific input parameters in order to determine the fundamental frequency component from the speech signals. The fundamental frequency component has been used for detection of voiced and non-voiced regions from speech signals. The instantaneous pitch frequency is computed using Hilbert transform of the fundamental frequency component corresponding to voiced regions of speech signals. The experimental results are shown on speech signals taken from Keele pitch extraction reference database. The experimental results obtained from the proposed method are compared with the other existing methods for determining pitch frequency from speech signals.