The Department of Electronic and Electrical Engineering is located on Thornton Science Park, a modern expressly-designed site that profits from a recently-completed multi-million pound renovation that has created a state-of-the-art teaching and research facility. The site was home to Shell UK’s exploration and research centre since the 1940s, and its takeover by the University heralded the opportunity to apply its legacy to the continuation of world-class innovation and research in the North West.

Recent Submissions

  • Alzheimer Brain Imaging Dataset Augmentation Using Wasserstein Generative Adversarial Network

    Ilyas, Kulsum; Hussain, B. Zahid; Andleeb, Ifrah; Aslam, Asra; Kanwal, Nadia; Ansari, Mohammad Samar; Aligarh Muslim University; University of Leeds; Keele University; University of Chester (Springer, 2024-02-25)
    Deep learning models have evolved to be very efficient and robust for several computer vision applications. To harness the benefits of state-of-the-art deep networks in the realm of disease detection and prediction, it is imperative that high-quality datasets be made available for the models to train on. This work recognizes the dearth of training data (both in terms of quality and quantity of images) for using such networks for the detection of Alzheimer’s disease. It is proposed to employ a Wasserstein Generative Adversarial Network (WGAN) for generating synthetic images for augmentation of an existing Alzheimer brain image dataset. It is shown that the proposed approach is indeed successful in generating high-quality images for inclusion in the Alzheimer image dataset potentially making the dataset more suited for training high-end models.
  • FireNet-Tiny: Very-Low Parameter Count High Performance Fire Detection Model

    Oyebanji, Olalekan J.; Oliver, Stefy; Ogonna, Chukwuka E.; Aslam, Asra; Ansari, Mohammad Samar; University of Chester; University of Leeds (Springer, 2024-02-25)
    In daily life, fire threats result in significant costs on the ecological, social, and economic levels. It is essential to outfit the assets with fire prevention systems due to the sharp rise in the frequency of fire mishaps. To prevent such mishaps, several studies have been conducted to develop optimal and potent fire detection models. While the earliest methods were thermal/chemical in nature, image processing was later applied for identification of fire. More recent methods have taken advantage of the significant advancements in deep learning models for computer vision. However, in order to maintain a suitable inference time (leading towards real-time detection) and parameter count, the majority of deep learning models have to make trade-offs between their detection speed and detection performance (accuracy/recall/precision). The very lightweight convolution neural network we offer in this paper is specifically designed for the fire detection use case. The proposed model can be embedded in real-time fire monitoring equipment and could also prove potentially useful for future fire monitoring methods such as unmanned aerial vehicles (drones). By further diminishing the trainable parameter count of the model, the fire detection results obtained using the proposed FireNet-Tiny significantly outperform the prior low parameter count models. When tested against FireNet dataset, FireNet-Tiny, which only comprises 261,922 parameters, was shown to have an overall accuracy of 95.75%. In comparison, FireNet-v2 provided 94.95% accuracy with 318,460 parameters.
  • FireNet-Micro: Compact Fire Detection Model with High Recall

    Ansari, Mohammad Samar; University of Chester; University of Leeds
    Fire occurrences and threats in everyday life incur substantial costs on ecological, economic, and even social levels. It is crucial to equip establishments with fire prevention systems due to the notable increase in fire incidents. Numerous studies have been conducted to develop efficient and optimal fire detection models in order to prevent such mishaps. Initially, thermal/chemical methods were used, but later, image processing techniques were also employed to identify fire occurrences. Recent approaches have capitalized on the advancements in deep learning models for computer vision. However, most deep learning models face a trade-off between detection speed and performance (accuracy/recall/precision) to maintain a reasonable inference time (for real-time applications) and parameter count. In this paper, we present a bespoke and highly lightweight convolutional neural network specifically designed for fire detection. This model can be integrated into real-time fire monitoring equipment and potentially applied in future methods suhc as CCTV surveillance cameras, traffic lights, and unmanned aerial vehicles (drones) for fire monitoring in futuristic smart city scenarios. Despite having significantly fewer trainable parameters, our customized model, FireNet-Micro, outperforms existing low-parameter-count models in fire detection. When evaluated on the FireNet dataset, FireNet-Micro, with only 171,234 parameters, achieved an impressive overall accuracy of 96.78%. In comparison, FireNet-v2 attained 94.95% accuracy with 318,460 parameters (which is almost double the parameter count of the proposed FireNet-Micro).
  • Deep Learning Based Lightweight Model for Brain Tumor Classification and Segmentation

    Andleeb, Ifrah; Hussain, B. Zahid; Ansari, Salik; Ansari, Mohammad Samar; Kanwal, Nadia; Aslam, Asra; Aligarh Muslim University; University of Chester; Keele University; University of Leeds (Springer, 2024-02-01)
    This paper presents two lightweight deep learning models for efficient detection and segmentation of brain tumors from MRI scans. A custom-made Convolutional Neural Network (CNN) is designed for identification of four different classes of brain tumors viz. Meningioma, Glioma, Pituitary brain tumor and normal (no tumor). Furthermore, another tailor-made lightweight model is presented for the segmentation of the tumor from the Magnetic Resonance Imaging (MRI) scans. The output of the segmentation model is the ‘mask’ depicting the tumor region. The overall performance in terms of detection accuracy, and segmentation accuracy, for the two models is found to be approximately 95% for both the cases individually. The proposed models are worthy additions to the existing literature on brain tumor classification and segmentation models due to their low-parameter count which make the models amenable for deployment on resource-constrained edge hardware.
  • Dual-WGAN Ensemble Model for Alzheimer’s Dataset Augmentation with Minority Class Boosting

    Ansari, Mohammad Samar; Ilyas, Kulsum; Aslam, Asra; University of Chester; Aligarh Muslim University; University of Leeds (IEEE, 2023-11-20)
    Deep learning models have become very efficient and robust for several computer vision applications. However, to harness the benefits of state-of-art deep networks in the realm of disease detection and prediction, it is crucial that high-quality datasets be made available for the models to train on. This work recognizes the lack of training data (both in terms of quality and quantity of images) for using such networks for the detection of Alzheimer’s Disease. To address this issue, a Wasserstein Generative Adversarial Network (WGAN) is proposed to generate synthetic images for augmentation of an existing Alzheimer brain image dataset. The proposed approach is successful in generating high-quality images for inclusion in the Alzheimer image dataset, potentially making the dataset more suitable for training high-end models. This paper presents a two-fold contribution: (i) a WGAN is first developed for augmenting the non-dominant class (i.e. Moderate Demented) of the Alzheimer image dataset to bring the sample count (for that class) at par with the other classes, and (ii) another lightweight WGAN is used to augment the entire dataset for increasing the sample counts for all classes.
  • Ferroelectric anomaly of perovskite layer structured Pb2+-doped Sr2Nb2O7 ceramics

    Liu, Lintao; Chen, Tao; Ouyang, Delai; Yue, Yajun; Yang, Bin; Yan, Haixue; Abrahams, Isaac; Fu, Zhengqian; Liang, Ruihong; Zhou, Zhiyong; et al. (Wiley, 2024-01-09)
    The spontaneous polarization of perovskite layer structured Sr2Nb2O7 ferroelectrics (FEs) is originated from mainly the oxygen octahedral rotations and partially the displacement of Sr2+ ions. However, there is FE anomaly of showing the typical characteristics of antiferroelectric (AFE)‐like behavior with double polarization–electric field hysteresis loops of Sr2Nb2O7 ceramics by Pb doping. Here, combinations of low frequency and sub‐terahertz band dielectric measurements under applied DC field reveal field‐induced transition from AFE to FE structure. Temperature dependence of dielectric constant suggested a second‐order phase transition near 215°C, which further supported by the lattice parameters and thermal expansion coefficient extracted from the variable temperature X‐ray diffraction in the heating and cooling processes. The selected area electron diffraction results show no new superlattice spots are observed along [1 0 0] zone axis related to c direction. We proposed a model based on octahedral tilting/rotation that accounted for the incommensurate lattice modulation in the c direction for Sr2Nb2O7 system by Pb doping. All the results show AFE‐like behavior be likely to origin from the electron structures of Pb2+ with a 6s2 lone pair. Our results gave us a new concept provide possibilities for the design of AFE‐like materials in layer structured compounds with super high FE Curie point.
  • The recent development of protection coordination schemes based on inverse of AC microgrid: A review

    Alasali, Feras; Mustafa, Haytham; Saidi, Abdelaziz Salah; El‐Naily, Naser; Abeid, Salima; Holderbaum, William; Omran, Emad; Saad, Saad M.; The Hashemite University; College of Electrical and Electronics Technology-Benghazi; King Khalid University; University of Chester; University of Salford; Suez Canal Authority (Wiley Open Access, 2023-12-06)
    Integration of distributed generation systems and diversity of microgrid operations led to a change in the structure of the power system. Due to this conversion, new challenges have arisen when employing traditional overcurrent protection schemes. As a consequence, non‐classical protection schemes have attracted significant attention in the last few years. Engineers and scholars have proposed different non‐standard methods to increase the power protection system and ensure the highly selectivity performance. Although the non‐standard characteristics and their requirements, in general, have been outlined and analyzed in the available literature, protection coordination based on voltage current–time inverse, as a branch of non‐standard optimization methods, has not yet been thoroughly discussed, compared, or debated in detail. To close this gap, this review introduces a broad overview of recent research and developments of the voltage current–time inverse based protection coordination. Focuses on assessing the potential advantages and disadvantages of related studies and provide a classification and analysis of these studies. The future trends and some recommendations have been included in this review for improving fault detection sensitivity and coordination reliability.
  • A Comparative Study of Accuracy in Major Adaptive Filters for Motion Artefact Removal in Sleep Apnea Tests

    Chen, Yongrui; Zheng, Yurui; Johnson, Sam; Wiffen, Richard; Yang, Bin; University of Chester; Passion for Life Healthcare (Springer, 2023-12-05)
    Sleep apnoea is probably the most common respiratory disorder, respiration and blood oxygen saturation (SpO2) are major concerns in sleep apnoea and are also the two main parameters checked by Polysomnography (PSG, the gold standard for diagnosing sleep apnoea). In this study, we used a simple, non-invasive monitoring system based on photoplethysmography (PPG) to continuously monitor SpO2 and heart rate (HR) for individuals at home. Various breathing experiments were conducted to investigate the relationship between SpO2, HR, and apnoea under different conditions, where two techniques (empirical formula and customized formula) for calculating SpO2 and two methods (resting HR and instantaneous HR) for assessing HR were compared. Various adaptive filters were implemented to compare the effectiveness in removing motion artefacts (MA) during the tests. This study fills the gap in the literature by comparing the performance of different adaptive filters on estimating SpO2 and HR during apnoea. The results showed that up-down finger motion introduced more MA than left-right motion, and the errors in SpO2 estimation were increased as the frequency of movement was increased; due to the low sampling frequency features of these tests, the insertion of adaptive filter increased the noise in the data instead of eliminating the MA for SpO2 estimation; the normal least mean squares (NLMS) filter is more effective in removing MA in HR estimation than other filters.
  • PES-g-BST/PEEK composites modified by surface grafting with high dielectric tunability

    Liu, Shuhang; Peng, Mingyu; Xu, Xin; Guo, Yiting; Wu, Sichen; Xu, Jie; Baxter, Harry; Yang, Bin; Gao, Feng; Northwestern Polytechnical University; Honghui Hospital of Xi’an Jiaotong University; University of Chester
    Ceramic/polymer composites have been widely utilized due to their outstanding dielectric and mechanical properties. The interfacial bonding between ceramic and organic phases has a significant effect on the properties of composites. Therefore, enhancing the interfacial bond strength has become a hot issue. In this work, a chemically modified PES-g-BST/PEEK composite was prepared via cold-pressing sintering. The chemical bond was constructed to connect one end of PES (polyether sulfone) with BST (barium strontium titanate) and the other end with PEEK (polyether ether ketone) by using 4,4′-Diaminodiphenylsulfone (DDS), which enhanced the interface combination between ceramic fillers and organic phase. The influence of the surface grafting method and the amount of DDS on the microstructure and dielectric characteristics of the PES-g-BST/PEEK composite was investigated, and the amount of DDS was optimized. The permittivity, dielectric tunable efficiency, and dielectric tunability of the composites were improved. A chemically modified PES-g-BST/PEEK composite with improved dielectric properties and high dielectric tunability was obtained. The dielectric tunability amounts to 38.16 % under a 7 kV/mm bias, accompanied by a low dielectric constant of 14.5 and a dielectric loss of 0.076 (1 kHz). This work provides a way of enhancing the interface bonding of ceramic/polymer composites to improve dielectric tunability.
  • Energy storage properties of samarium-doped bismuth sodium titanate-based lead-free ceramics

    Tang, Xuyao; Hu, Zimeng; Koval, Vladimir; Yang, Bin; Smith, Graham C.; Yan, Haixue; University of Chester (Elsevier, 2023-08-18)
    Due to worldwide environmental regulations, lead-free relaxors, namely Bi0.5Na0.5TiO3–6BaTiO3 (BNT-6BT) are being extensively studied as an alternative candidate for energy storage applications. Here, Sm was introduced at different A sites of the relaxor system; specifically, the Sm-doped BNT-6BT system was designed to replace Bi (BNT-Bi), Na (BNT-Na), and both the Bi and Na ions (BNT-BiNa) by Sm ions. It was found that the BNT-Bi sample possesses high piezoelectricity (d33=117.3 pC N-1), whereas the BNT-Na and BNT-BiNa ceramics show exceptionally high values of the energy storage density and efficiency. To define the energy storage performance, a new concept based on determining the recoverable energy storage intensity is proposed in the present work. This allows bypassing the high applied electric fields in determining the value of the energy storage density. An ultrahigh recoverable energy storage density (4.41 J cm-3), excellent energy storage efficiency (83.96%) and superhigh recoverable energy storage intensity (19.17×10-3 J kV-1 cm-2) were achieved in the BNT-BiNa ceramics simultaneously. Furthermore, the energy storage characteristics exhibit an excellent stability over a wide temperature range from 25 °C to 150 °C. Thus, the developed Sm-doped BNT-6BT ceramics show great potential for piezoelectric and high-power energy storage applications.
  • Broadband Dielectric Characterization of Carbon Black-Reinforced Natural Rubber

    Huang, Menglong; Tunnicliffe, Lewis B.; Liao, Shibai; Yang, Bin; Yan, Haixue; Busfield, James J. C.; Queen Mary University of London; University of Chester (Rubber Division, ACS, 2023-08-28)
    Natural rubber compounds reinforced with two different carbon blacks (N134 and N330) at various concentrations were characterized using very broadband dielectric spectroscopy from around 0.1 Hz to 0.3 THz using four different impedance and network analysis technologies. Percolation behavior was observed when the testing electrical frequency was below a certain range, which can be linked to the presence of percolated carbon black networks. When above a critical frequency level, the real part of AC conductivity or the permittivity tended to have a simple exponential relationship with the volume fraction of carbon black rather than a percolation-like behavior with the carbon black volume fraction and was no longer sensitive to carbon black networks. The AC conductivity derived via complex impedance was also strongly influenced by the choice of calculation model when the material was around or below the percolation threshold.
  • High quality of LiMg0.9Zn0.06Ni0.04PO4-TiO2 microwave ceramic and its application for 5G dielectric waveguide bandpass filter

    Chen, Long; Liu, Huan; Jiang, Yu; Li, Shuai; Luo, Xinjiang; You, Bin; Li, Aihua; Hu, Yuanyun; Baxter, Harry; Yang, Bin; et al. (Elsevier, 2023-01-30)
    A new microwave dielectric composite ceramic (1-x) wt%LiMg0.9Zn0.06Ni0.04PO4-x wt%TiO2 (0≤x≤18wt%) was synthesized in the low sintering temperature range of 850 oC -975 oC by the solid-state reaction method. The XRD diffractogram confirmed the coexistence of LiMg0.9Zn0.06Ni0.04PO4 and TiO2, and no second phase was detected. Due to their opposite τf values, the τf values of LiMg0.9Zn0.06Ni0.04PO4 solid solution ceramics were adjusted by adding different volume fractions of TiO2. The composite ceramic with x=15wt% sintered at 975 oC shows desirable microwave dielectric properties of εr~10.30, Q×f ~ 58,400GHz (tanδ=2.093×10-4), and τf ~ +4.04ppm/ oC (at ~12GHz). Based on the 85wt% LiMg0.9Zn0.06Ni0.04PO4-15wt% TiO2 ceramics, a fourth order dielectric waveguide filter filled with composite ceramics is designed and packaged for the RF front-end of the China Mobile 5G base station. The test results show that the center frequency of the filter is 4.65GHz, the working bandwidth is 300MHz, the filter has a good out of band rejection capability, and the insertion loss in band is 0.18dB. Combined with the filter test data and material performance, the designed dielectric waveguide filter meets the communication requirements of the base station, and the higher Q×f microwave dielectric ceramic material can achieve low insertion loss and excellent frequency selection performance of the filter.
  • (Ba0.6Sr0.4)TiO3/PEEK composites modified by Polyethersulfone with low dielectric constant and high dielectric tunability under DC bias

    Yang, Bin; Liu, Shuhang; Guo, Yiting; Hu, Guoxin; Wu, Sichen; Xu, Jie; Chen, Jianxin; Bulejak, Weronika; Baxter, Harry; Kong, Jie; et al. (Elsevier, 2023-01-19)
    Ceramic/polyetheretherketone (PEEK) composites show a wide range of applications and have attracted extensive interest in the scientific community due to their outstanding dielectric and mechanical characteristics. However, the interface connection between the ceramic and PEEK is a vital issue that must be addressed to improve their physical and electrical properties. In this work, the polyethersulfone resin (PES) was selected as interface modifier between barium strontium titanate (Ba0.6Sr0.4TiO3, BST) and PEEK. Cold-pressing sintering was used to create BST/PEEK materials with superior dielectric frequency stability and dielectric tunability. The effects of PES content on the morphology and dielectric characteristics of PES modified BST/PEEK materials were investigated. The results showed PES could improve the dispersion of BST particles in polymer. The dielectric constant, dielectric tunability, and breakdown strength increased first, then reduced as PES content increased. The composite had the most homogeneous microstructure and the best dielectric properties when the PES content was 7.5vol%. The frequency dispersion factor F(x) was much smaller than that of other ceramic/polymer composites reported. In addition, the dielectric tunability of the composites could reach a relatively high level (34.18%) while the dielectric constant was as low as 14. The dielectric tunable efficiency (TuE) was proposed to evaluate the property of low dielectric constant and high dielectric tunability under DC bias. The TuE of PES modified BST/PEEK composites show the highest value comparing with reported dielectric tunable composites. This research laid the path for the development of a novel ceramic/polymer composite with good interface bonding and high dielectric tunability.
  • THz probing ferroelectric domain wall dynamics

    Yang, Bin; Zhang, Man; Yan, Haixue; University of Chester; Queen Mary University of London (IEEE, 2022-12-27)
    This work uses THz time domain spectroscopy (THz-TDS) to detect the dynamics of domain walls in a Aurivillius phase ferroelectric ceramic, Ca 0.96 Rb 0.02 Ce 0.02 Bi 2 Nb 2 O 9 . Results show that ferroelectric domain walls are active at the THz band, with lower dielectric permittivity compared with that of the domain. This work has verified that it is feasible to use domain wall engineering method to optimize properties of ferroelectrics at the THz band, which help create new applications for ferroelectric materials at THz frequency.
  • New methodology to reduce power by using smart street lighting system

    Al-khaykan, Ameer; Aziz, Ali Saleh; Al-Kharsan, Ibrahim H.; Counsell, John M.; Mustaqbal University College; Al-Hussain University College; The Islamic University; University of Chester (De Gruyter, 2022-12-08)
    One of most important things now is to create smart street and smart lighting system to save enormous electrical energy. Especially Iraq is suffering shortage of electrical energy generation up to 45%. Because of this, Iraq needs to save a lot of electrical energy in the entire country so as to meet the electrical demand and reduce the large amount of CO2 emission. However, this work presents a very unique and economic control lighting system (CLS) for main streets and sidewalks, which can control the lighting system to give sufficient illumination to the drivers and the pedestrians simultaneously. And at the same time, the CLS system can reduce a lot of electrical energy consumption and the CO2 emissions together. However, by using these smart systems with the exciting illumination source in the streets, the CLS can minimize the electrical energy consumed for the lighting at the main roads and the footpath by about 60% and can use the surplus energies to fill the shortage of electricity in the country. Also, this system will increase the lifetime of the lighting system which means further decrease in cost. Finally, this work presents new type of illumination source, high-intensity discharge (HID), which can reduce the electrical consumption much more by up to 90%, when using the CLS with HID.
  • Rare earth ion-doped Y2.95R0.05MgAl3SiO12 (R = Yb, Y, Dy, Eu, Sm) garnet-type microwave ceramics for 5G application

    Ye, Zijun; Jiang, Yu; Mao, Minmin; Xiu, Zhiyu; Chi, Mengjiao; Wu, Guofa; Liu, Bing; Wang, Dawei; Yang, Bin; Song, Kaixin; et al. (MDPI, 2022-11-11)
    In this work, Y2.95R0.05MgAl3SiO12 (R=Yb, Y, Dy, Eu, Sm) microwave single-phase dielectric ce-ramics were successfully prepared via conventional ceramic technology by doping a series of rare earth elements with different ionic radius (Yb, Y, Dy, Eu, Sm) for the first time. The effects of A site occupied by rare earth elements on the microwave dielectric properties of Y2.95R0.05MgAl3SiO12 were studied by crystal structure refinement, scanning electron microscope (SEM), bond valence theory, P-V-L theory and infrared reflection spectroscopy. It was found that the ionicity of Y-O bond, the lattice energy, the bond energy and bond valance of Al(Tet)-O bond had important effects on microwave dielectric properties. Particularly, the optimum microwave dielectric properties were obtained for Y2.95Dy0.05MgAl3SiO12 sintered at 1575 °C for 6 h, with εr = 9.68, Q×f = 68,866 GHz, and τf = -35.8 ppm/°C, displaying its potential prospect in the 5G communication.
  • Terahertz Faraday rotation of SrFe12O19 hexaferrites enhanced by Nb-doping

    Hu, Zimeng; Stenning, Gavin B. G.; Koval, Vladimir; Wu, Jiyue; Yang, Bin; Leavesley, Alisa; Wylde, Richard; Reece, Michael J.; Jia, Chenglong; Yan, Haixue; et al. (American Chemical Society, 2022-10-04)
    The magneto-optical and dielectric behaviour of M-type hexaferrites as permanent magnets in the THz band are essential for potential applications like microwave absorbers and antennas, while are rarely reported recent years. In this work, single-phase SrFe12-xNbxO19 hexaferrite ceramics were prepared by conventional solid state sintering method. Temperature-dependent of dielectric parameters were investigated here to search the relationship between dielectric response and magnetic phase transition. The saturated magnetization increases by nearly 12% while the coercive field decreases by 30% in the x = 0.03 composition compared to that of the x = 0.00 sample. Besides, Nb substitution improves the magneto-optical behaviour in the THz band by comparing the Faraday rotation parameter from 0.75 (x = 0.00) to 1.30 (x = 0.03). The changes in the magnetic properties are explained by a composition-driven increase of the net magnetic moment and enhanced ferromagnetic exchange coupling. The substitution of donor dopant Nb on the Fe site is a feasible way to obtain multifunctional M-type hexaferrites, as preferred candidates for permanent magnets, sensors and other electronic devices.
  • FinFET-based non-linear analog signal processing modules

    Sharma, Vipin Kumar; Ansari, Mohammad Samar; Parveen, Tahira; Aligarh Muslim University; University of Chester (Elsevier, 2022-11-08)
    FinFETs exhibit far superior transistor characteristics (better gate control and a lower sub-threshold slope) as compared to the standard MOSFETs, this paper first employs the FinFETs in the design of an operational transconductance amplifier (OTA). The FinFET-based OTA offers a linearity range and - 3 dB bandwidth of 300 mV and 631.81 GHz, respectively. Further, the non-linear applications of the proposed OTA, viz. voltage divider, memristor emulator, and a memristive neuron, are presented. The proposed analog voltage divider circuit contains one OTA and two external N-FinFETs. The maximum bandwidth obtained for the voltage divider is 217.54 GHz. The memristor emulator contains one OTA, two external N-FinFETs, and one grounded capacitor. The proposed emulator circuit follows the signature characteristics of the actual memristor device. The frequency response characteristics of the proposed emulator circuit depict a bandwidth of 22.7 GHz. The proposed emulator shows non-volatile as well as electronically tunable features. Next, Monte-Carlo simulation analysis has been performed on the proposed circuits in order to observe the effects of statistical variation in different operating conditions. Furthermore, we propose a FinFET-based passive memristive neuron model using a memristor emulator circuit. The proposed neuron circuit follows a tangent hyperbolic activation function. All the proposed circuits are suitable for monolithic implementation. The proposed circuits are verified using 20 nm FinFET technology. The simulation results obtained using HSPICE agree well with the theoretical analysis.
  • A Novel Double-Threshold Neural Classifier for Non-Linearly Separable Applications

    Kashif, Mohd; Rahman, Syed Atiqur; Ansari, Mohammad Samar; Aligarh Muslim University; University of Chester (IEEE, 2023-03-28)
    Classification of data finds applications in various engineering and scientific problems. When real-time operation is desired, hardware solutions tend to be more amenable as compared to algorithmic/heuristic solutions. This paper presents a novel current-mode dual-threshold neuron designed and implemented at 32nm CMOS technology node. Subsequently, a current-mode double-threshold classifier is presented which is capable of classifying input patterns of non-linearly separable problems. Thereafter, application of the current-mode dual-threshold neuron in the realization of the XOR function using only a single neural unit is discussed. The proposed neuron as well as both the applications discussed are capable of operating from sub-1V power supplies. Computer simulations using HSPICE yield promising results with the values of delay and power consumption estimated to be lower than existing circuits.
  • Wasserstein GAN based Chest X-Ray Dataset Augmentation for Deep Learning Models: COVID-19 Detection Use-Case

    Hussain, B. Zahid; Andleeb, Ifrah; Ansari, Mohammad Samar; Joshi, Amit Mahesh; Kanwal, Nadia; Aligarh Muslim University; University of Chester; Malaviya National Institute of Technology Jaipur; Keele University (IEEE, 2022-09-08)
    The novel coronavirus infection (COVID-19) is still continuing to be a concern for the entire globe. Since early detection of COVID-19 is of particular importance, there have been multiple research efforts to supplement the current standard RT-PCR tests. Several deep learning models, with varying effectiveness, using Chest X-Ray images for such diagnosis have also been proposed. While some of the models are quite promising, there still remains a dearth of training data for such deep learning models. The present paper attempts to provide a viable solution to the problem of data deficiency in COVID-19 CXR images. We show that the use of a Wasserstein Generative Adversarial Network (WGAN) could lead to an effective and lightweight solution. It is demonstrated that the WGAN generated images are at par with the original images using inference tests on an already proposed COVID-19 detection model.

View more