BITS Faculty Publications

Permanent URI for this communityhttp://localhost:4000/handle/123456789/1867

Browse

Now showing 1 - 20 of 38

Adaptive artificial neural network based control strategy for shunt active power filter
(IEEE, 2016) Ajmera, Pawan K.; Bansal, Hari Om
Shunt active power filter (SAPF) is used to mitigate the current harmonics and to improve the power factor. In this paper, Adaptive linear-neuron (ADALINE) based phase lock loop (PLL) controlling scheme is presented for SAPF. ADALINE networks estimate the fundamental supply frequency. This scheme detects the phase information of the supply voltage and also used for parallel computing as it provides faster convergence. This algorithm is trained by least-mean squares (LMS) rule which offers low computational burden on the system. In this work, ADALINE is tuned using particle swarm optimization (PSO) technique to improve the dynamic performance of the system. The results obtained are compared with conventional PLL control technique and are found to be significantly better. The performance of the proposed ADALINE based control algorithm is validated using MATLAB/Simulink.
Adaptive Artificial Neural Network Based Control Strategy for Shunt Active Power Filter” IEEE International Conference on Electrical Power and Energy Systems (ICEPES), December 14-16, 2016, MANIT Bhopal, India.
(IEEE, 2016) Bansal, Hari Om; Ajmera, Pawan K.
Shunt active power filter (SAPF) is used to mitigate the current harmonics and to improve the power factor. In this paper, Adaptive linear-neuron (ADALINE) based phase lock loop (PLL) controlling scheme is presented for SAPF. ADALINE networks estimate the fundamental supply frequency. This scheme detects the phase information of the supply voltage and also used for parallel computing as it provides faster convergence. This algorithm is trained by least-mean squares (LMS) rule which offers low computational burden on the system. In this work, ADALINE is tuned using particle swarm optimization (PSO) technique to improve the dynamic performance of the system. The results obtained are compared with conventional PLL control technique and are found to be significantly better. The performance of the proposed ADALINE based control algorithm is validated using MATLAB/Simulink.
Adaptive Quality Enhancement Fingerprint Analysis
(IEEE, 2020) Ajmera, Pawan K.
Poor quality of the fingerprint image prevents accurate recognition as the employed methods are largely dependent on the fingerprint image quality. Algorithms will be better suited to detect fingerprint images if they are adapted according to their quality classes. In this paper, a class adaptive fingerprint enhancement algorithm is presented, classes dry, good and wet are assigned and further image processing is carried out. Features such as mean, variance, moisture index, Ridge Valley Area Uniformity (RVAU) are extracted from the ROI images. There are two stages of fingerprint quality enhancements which include the quality preprocessing (QP) and the enhancement stage. Support Vector Machine (SVM) algorithm is used to classify the images. Further, comparison scores are calculated by comparing the given image with the database of the minutiae using the minutiae matching technique. Experimentation is carried out on the FVC fingerprint database. A comparative analysis of the fuzzy C-means based clustering and mean based clustering is also experimented.
AMaizeD: An End to End Pipeline for Automatic Maize Disease Detection
(IEEE, 2023) Ajmera, Pawan K.
This research paper presents AMaizeD: An End to End Pipeline for Automatic Maize Disease Detection, an automated framework for early detection of diseases in maize crops using multispectral imagery obtained from drones. We also develop a custom hand-collected dataset focusing specifically on maize crops was meticulously gathered by expert researchers and agronomists. The dataset encompasses a diverse range of maize varieties, cultivation practices, and environmental conditions, capturing various stages of maize growth and disease progression. By leveraging multispectral imagery, the framework benefits from improved spectral resolution and increased sensitivity to subtle changes in plant health. The proposed framework employs a combination of convolutional neural networks (CNNs) as feature extractors and segmentation techniques to identify both the maize plants and their associated diseases. Experimental results demonstrate the effectiveness of the framework in detecting a range of maize diseases, including common rust, grey leaf spot and leaf blight. The framework achieves state-of-the-art performance on the custom hand-collected dataset and contributes to the field of automated disease detection in agriculture, offering a practical solution for early identification of diseases in maize crops using advanced machine learning techniques and deep learning architectures.
Analytical study on users’ awareness and acceptability towards adoption of multimodal biometrics (MMB) mechanism in online transactions: a two-stage SEM-ANN approach
(Springer, 2022-09) Ajmera, Pawan K.
The study analyses user awareness of multimodal biometrics and its acceptability for online transactions in the current dynamic world. The study was performed on the five underlying perspectives: User Acceptability, Cognizant Factors towards Biometrics, Technological factors, Perceptional Factors (Fingerprints, Iris, Face Recognition and Voice) and Data Privacy Factors. A questionnaire was prepared and circulated to the 530 biometrics users; on that basis, the corresponding answer was obtained for analysis. SEM is first employed to gauge the research model and test the prominent hypothesized predictors, which are then used as inputs in the neural network to evaluate the relative significance of each predictor variable. By considering the standardized significance of the feed-for-back-propagation of ANN algorithms, the study found a significant effect of DPF_3 (93%), DPF_2 (50%) and DPF_4 (34%) on the adoption of MMB. In the Perceptional construct, PRF_2 (49%) and PRF_3 (33%) was relatively the most important predictor, whereas, in User Acceptability, UAC_2 (37%), UAC_3 & UAC_5 (41%) was vital to be considered. Only one item, TCF_2 (35%), from Technological Factors, followed by Cognizant factors, i.e., CFG_1 (33%), confirmed the best fit model to adopt MMB. The research is a novel effort when compared to past studies as it considered cognizant and perceptual factors in the proposed model, thereby expanding the analytical outlook of MMB literature. Thus, the study also explored several new and valuable practical implications for adopting multimodal instruments of biometrics along with certain limitations.
ANN based approach for selective detection of breath acetone by using hybrid GO-FET sensor array
(IOP, 2022-04) Ajmera, Pawan K.; Hazra, Arnab
This research used hybrid graphene oxide (GO) field effect transistors (FETs) based sensor array to design an electronic nose (e-nose) for identifying exhaled breath acetone to diagnose diabetes mellitus through noninvasive route. Six back gated FET sensors were fabricated with hybrid channel of GO, WO3 and noble metals (Au, Pd and Pt) nanoparticles. The experiment was carried out by using four distinct forms of synthetic breath, each with a different level of interference. Linear discriminant analysis (LDA) and artificial neural networks (ANN) were utilized to classify and analyze the sensor response vector. In contrast, partial least square (PLS) and multiple linear regression (MLR) were used to evaluate the exact acetone concentration in synthetic breath. First, LDA was used to lower the dimensionality of the response vector, which was then provided as an input to the ANN model. ANN was performed with ten perceptrons model in the hidden layer and highest accuracy of 99.1% was achieved. Additionally, by using the loading plot of PLS, three sensors (Pt/WO3/GO, Pd/WO3/GO, and WO3/GO) had the ample use to predict the concentration of breath acetone. Moreover, the MLR approach with correlation coefficient (R2) of 0.9572 and root mean square error (RMSE) of 5.63% were used for obtaining the exact concentration of acetone. Consequently, e-nose with matrix of hybrid GO-FET sensors and pattern recognition algorithms (LDA, ANN, PLS and MLR) exhibited considerable ability in selective detection of acetone in synthetic breath.
APiCroDD: Automated Pipeline for Crop Disease Detection
(Springer, 2024) Ajmera, Pawan K.
This research paper proposes APiCroDD: automated pipeline for crop disease detection, an automated framework for early detection of plant diseases using multispectral imagery from drones. Current frameworks for disease detection are labor and time-consuming. They do not leverage the richness of multispectral imagery for feature extraction and perform vanilla manipulation of agriculture indices. Our framework comprises two stages: data acquisition and disease identification. We find that the use of multispectral imagery in the proposed framework provides several advantages over traditional RGB imagery, including better spectral resolution and increased sensitivity to subtle changes in plant health. The multispectral data enables the identification of specific spectral bands associated with diseased regions of the plant, improving the accuracy of disease detection. The proposed framework utilizes a combination of CNNs and segmentation techniques to identify the plant and its disease. Experimental results demonstrate that the proposed framework using EfficientNet is highly effective in identifying a range of plant diseases achieving state-of-the-art performance on manually collected dataset and validated on the PlantVillage dataset.
Audio classification using braided convolutional neural networks
(IET, 2020-09) Ajmera, Pawan K.
Convolutional neural networks (CNNs) work surprisingly well and have helped drastically enhance the state-of-the-art techniques in the domain of image classification. The unprecedented success motivated the application of CNNs to the domain of auditory data. Recent publications suggest hidden Markov models and deep neural networks for audio classification. This study aims to achieve audio classification by representing audio as spectrogram images and then use a CNN-based architecture for classification. This study presents an innovative strategy for a CNN-based neural architecture that learns a sparse representation imitating the receptive neurons in the primary auditory cortex in mammals. The feasibility of the proposed CNN-based neural architecture is assessed for audio classification tasks on standard benchmark datasets such as Google Speech Commands datasets (GSCv1 and GSCv2) and the UrbanSound8K dataset (US8K). The proposed CNN architecture, referred to as braided convolutional neural network, achieves 97.15, 95 and 91.9% average recognition accuracy on GSCv1, GSCv2 and US8 K datasets, respectively, outperforming other deep learning architectures.
Automated Person Authentication Using Face, Iris and Ear Multimodal Biometric Fusion Chapter Automated Person Authentication Using Face, Iris and Ear Multimodal Biometric Fusion
(CRC, 2020) Ajmera, Pawan K.
With the rapidly growing fields of Computer Vision and Artificial Intelligence, automatic person identification is very crucial and important task associated with these technologies. Integrating information which is originating from different sources is one of the major factors for multimodal biometric system design. As multimodal biometrics basically involves more than two modalities so in this proposed work, the implementation of personal identification is performed by fusing face, iris and ear biometric modalities using feature level fusion approach
Comparative study of License Plate Recognition
(IJERT, 2014-03) Ajmera, Pawan K.
License Plate Recognition (LPR) is the most interesting and challenging area of research due to its importance to a wide range of commercial application. It is known that the number plates differ in shape and size. License plate recognition (LPR) is an image processing technology used to extract vehicle information from their license plates. There are three modules in license plate recognition system. 1) Detection 2) Character segmentation 3) Character recognition. There are many techniques for license plate detection, each having its own advantages and drawbacks. The basic step of license plate detection is localization of number plate. This paper compares morphological method, histogram based method and mathematical morphology methods of license plate detection. Character segmentation is done by using connected component and thresholding method. Character recognition is done by template matching method. The aim of this paper is to study and evaluates some of the most important license plate recognition algorithms
Convolutional Neural Network-Based Human Identification Using Outer Ear Images
(Springer, 2018-10) Ajmera, Pawan K.; Sinha, Yash
This paper presents a deep learning approach for ear localization and recognition. The comparable complexity between human outer ear and face in terms of its uniqueness and permanence has increased interest in the use of ear as a biometric. But similar to face recognition, it poses challenges such as illumination, contrast, rotation, scale, and pose variation. Most of the techniques used for ear biometric authentication are based on traditional image processing techniques or handcrafted ensemble features. Owing to extensive work in the field of computer vision using convolutional neural networks (CNNs) and histogram of oriented gradients (HOG), the feasibility of deep neural networks (DNNs) in the field of ear biometrics has been explored in this research paper. A framework for ear localization and recognition is proposed that aims to reduce the pipeline for a biometric recognition system. The proposed framework uses HOG with support vector machines (SVMs) for ear localization and CNN for ear recognition. CNNs combine feature extraction and ear recognition tasks into one network with an aim to resolve issues such as variations in illumination, contrast, rotation, scale, and pose. The feasibility of the proposed technique has been evaluated on USTB III database. This work demonstrates 97.9% average recognition accuracy using CNNs without any image preprocessing, which shows that the proposed approach is promising in the field of biometric recognition.
Encryption of Gray Images using Scalable Codes
(IJERT, 2014-04) Ajmera, Pawan K.
In this paper we have used a method for encryption of images using scalable coding. In encryption, a pseudorandom number derived from a secret key is used to encrypt the original pixel values. In encoding block, encrypted data is decomposed into a down sampled sub image and several data sets with a multiple-resolution construction. Thus an encoder quantizes the sub image and calculates the Hadamard coefficients of each data set that will reduce the data amount. Then, this quantized sub image and Hadamard coefficients are regarded as a set of bitstreams. At the receiver end, while decryption of sub image will give the rough information of the original content, the quantized coefficients can be decrypted to regenerate the detailed content with an iteratively updating procedure. At the output we will first obtain a lossy version of image and images of higher resolution can be obtained when more and more bitstreams are received. PSNR of reconstructed images is calculated for all scales. When more number of decomposition levels are used to decompose encrypted data, PSNR performance is good.
Face Recognition using Local Texture Descriptor
(IJERT, 2014-04) Ajmera, Pawan K.
A face recognition system is a computer application for automatically identifying a person from a digital image. Recognition of face in uncontrolled lightening situations is one of the most important bottlenecks for practical face recognition systems. This paper addresses the problem of illumination effects on Face recognition and work for an approach to reduce their effect on recognition performance. For this following methods are used: (i) simple and efficient preprocessing chain that eliminates most of the effects of changing illumination while still preserving the essential appearance details that are needed for recognition; (ii) Local Binary Pattern (LBP) texture descriptor which labels the pixels of an image and gives output as a histogram of image; and (iii) principle component analysis (PCA) feature extraction algorithm is used to improve robustness. The proposed method is tested on ORL face database. The crux of the work lies in optimizing Euclidean distance classifier for recognition of face.
Field-Assisted Sensitivity Amplification in a Noble Metal Nanoparticle Decorated WO3/GO Hybrid FET-Based Multisensory Array for Selective Detection of Breath Acetone
(Wiley, 2021-12) Ajmera, Pawan K.; Singhal, Rahul; Hazra, Arnab
The current work shows noble metal (Au, Pd and Pt) nanoparticles, WO3 and graphene oxide based ternary hybrid field effect transistor structured sensor array for the selective detection of breath acetone. Accurate acetone selectivity was achieved by using phase space entire based feature extraction techniques. Field-assisted sensitivity amplification was responsible for the detection of ppb to lower ppm level detection of acetone.
Fractional Fourier transform based features for speaker recognition using support vector machine
(Elsevier, 2013-02) Ajmera, Pawan K.
This paper presents a text-independent speaker recognition technique in which the conventional Fourier transform in Mel-Frequency Cepstral Coefficient (MFCC) front-end is substituted by fractional Fourier transform. Support Vector Machine (SVM) maps these input features into a high-dimensional space to separate classes by a hyperplane with enhanced discrimination capability. SVM based on mean-squared error classifier produces more accurate system. The Fractional Fourier Transform (FrFT) reveals the mixed time and frequency components of the signal. Modelling of speech signals as mixed time and frequency signals represents better production and perception speech characteristics. Processing of time-varying signals in fractional Fourier domain allows us to estimate the signal with least Mean Square Error (MSE) making the technique robust against additive noise compared to Fourier domain maintaining same computational complexity. The feasibility of the proposed technique has been tested experimentally using Texas Instruments and Massachusetts Institute of Technology (TIMIT) and Shri Guru Gobind Singhji (SGGS) databases. The experimental results show the superiority of the proposed method.
A multi-modal and multi-algorithmic biometric system combining iris and face
(IEEE, 2015) Ajmera, Pawan K.
In this paper, we have developed an algorithm which combines features from human iris and face for person verification. Iris recognition is one of the most accurate biometric modalities having verification results close to 98%. On the other hand, face is one of the most widely used biometric features because of its ease of capture. We have adapted score level fusion strategy for our system. However, in addition to this, we are using two different features for face: Gabor filters based and Local Binary Patterns (LBP) based. The iris features are extracted using Daugman's Gabor filters based approach. Using this information, we have developed a multi-modal (combining iris and face), multi-algorithmic (using two different algorithms for feature extraction from face) biometric system. With this system, we achieved more than 85% improvement in the verification performance in terms of Equal Error Rate as compared to the uni-biometrics based system.
Multi-modal biometric fusion based continuous user authentication for E-proctoring using hybrid LCNN-Salp swarm optimization
(Springer, 2021-11) Ajmera, Pawan K.
In Covid 19, pandemic remote proctoring of the employee or human being is evolved as a big challenge for the information retrieval process. On the other side, memory-based system access authentication is becoming outdated and less preferred for live applications, especially where data security and customer privacy are crucial. Multi-modal authentication has outperformed the unimodal process with high accuracy and improved security in the user authentication field. Multi-modal biometric verification includes user attributes such as keystrokes, iris, speech, face, etc. For real-time execution of multi-modal biometric fusion-based live tracking for compatible applications. The study proposes an efficient continuous biometric user authentication system for a new challenge of pandemic time, a live online authentication of the evaluation process (CBUA-OE). The proposed CBUA-OE system can address the challenges associated with live proctoring and is also compatible with real-time implementation, deployment of authentication systems. The modified wolf optimization algorithm and CUBA-OE's optimal feature fusion algorithm give an edge over the other contemporary methods and make it more robust. In modern forms of authentication, the classification stage affects the overall outcome of the system, and the model's performance is also a factor of varying quality of datasets. In contrast, a hybrid LCNN-Salp swarm optimization-based classifier is more efficient and consistent in continuous user authentication. Here the performance of the proposed hybrid LCNN-Salp swarm optimization classifier is analyzed with different standard datasets. The results are compared with the existing state-of-art classifiers regarding the accuracy, precision, recall, and F-measure. This projected work is novel in terms of usability factors and scalability to live tracking systems.
Multi-view Feature Learning Based on Texture Description for Palm-Print Recognition
(Springer, 2023-09) Ajmera, Pawan K.
Biometric is the science of validating an individual’s identity while using behavioral and physiological characteristics. In unconstrained scenario, contactless palm-print recognition leads to better recognition accuracy of individuals. Most of the existing texture descriptors are fail to learn stable and discriminative features from palm-print images. The paper presents a multi-view feature learning method based on texture description for palm-print recognition. The multi-view features are simultaneously extracted by two complementary operators. We also learn how to use feature mapping to convert multi-view data into hash codes. Experiments are carried out on palm-print databases captured using a variety of devices and acquisition methods. We demonstrate that the proposed method has superior performance compared to the current methods.
Multimodal Multilevel Fusion of Face Ear Iris with Multiple Classifications
(Springer, 2020-07) Ajmera, Pawan K.
With the advancement in the computational efficiency, there is also simultaneous increase in many efficient and secure biometric systems that are capable for the use of multiple sources of access authorization. Single biometric systems are inefficient and less secure which give rise to the advancement of multimodal biometric systems. Also, fusion of many biometric modalities is high area of interest, and here, many methods are deployed for the fusion of biometric data. Multimodal biometric system provides many evidences for the same person. In this paper, the design of multimodal biometrics based on face, ear, and iris modalities with multilevel fusion-based approach is preferred. In the presented work with multilevel multimodal fusion, 95.09% accuracy has been obtained which is better than highest unimodal accuracy; in this case, it is iris 94.06%. The obtained results are better than similar multimodal fusion-based model with single classifiers such as RNN with 90.58% accuracy and KNN classifier with 91.22% accuracy. So, in this work multilevel fusion of (i) different unimodal methods with (ii) feature level fusion of multiple traits has been proposed for person identification.
Multiresolution Feature Based Subspace Analysis for Fingerprint Recognition
(International Journal of Computer Applications, 2010-02) Ajmera, Pawan K.
The image intensity surface in an ideal fingerprint image contains a limited range of spatial frequencies, and mutually distinct textures differ significantly in their dominant frequencies. This paper presents a multiresolution feature based subspace technique for fingerprint recognition. The technique computes the core point of fingerprint and crops the image to predefined size. The multiresolution features of aligned fingerprint are computed using 2-D discrete wavelet transform. LL component in wavelet decomposition is concatenated to form the fingerprint feature. Principal component analysis is performed on these features to extract the features with reduced dimensionality. The algorithm is effective and efficient in extracting the features. It is also robust to noise. Experimental results using the FVC2002 and Bologna databases show the feasibility of the proposed method.