The latest WHO report showed that the number of malaria cases climbed to 219 million last year, two million higher than last year. The global efforts to fight malaria have hit a plateau and the most significant underlying reason is international funding has declined. Malaria, which is spread to people through the bites of infected female mosquitoes, occurs in 91 countries but about 90% of the cases and deaths are in sub-Saharan Africa. The disease killed 4,35,000 people last year, the majority of them children under five in Africa. AI-backed technology has revolutionized malaria detection in some regions of Africa and the future impact of such work can be revolutionary. The malaria Cell Image Data-set is taken from the official NIH Website NIH data. The aim of the collection of the dataset was to reduce the burden for microscopists in resource-constrained regions and improve diagnostic accuracy using an AI-based algorithm to detect and segment the red blood cells. The goal of this work is to show that the state of the art accuracy can be obtained even by using 2 layer convolution network and show a new baseline in Malaria detection efforts using AI.

In recent years, the field of digital image watermarking has garnered considerable attention due to its critical role in safeguarding broadcasting media. This technique serves as a powerful tool for ensuring the authenticity of content, monitoring media broadcasts, managing duplication control, and tackling a variety of related issues. Researchers have extensively examined digital image watermarking to address these challenges, making it a pivotal area of study. This paper delves into the concept of digital image watermarking, starting with an overview of its fundamental model. It then explores the primary requirements for effective watermarking systems and highlights their diverse applications. Additionally, the paper reviews key techniques and algorithms employed in watermarking, analyzes potential threats and attacks that such systems may face, and outlines the methods used to evaluate the performance and reliability of these systems.

The demand for efficient large integer polynomial multiplications in present day crypto-systems is the need of the hour. Karatsuba-like multiplication is one of the most efficient multiplication algorithm discussed in this work. However, this algorithm is mostly practically avoided due to the presence of complex sub-multiplications at the intermediate steps of computation. Thus, efforts has been made to implement two-term Karatsuba Multiplication (Method-I & Method-II), i.e., TTKM-I and TTKM-II in terms of speed and hardware utilization. The overall performance of the proposed design methods are also noted by calculating Area-Time-Product (ATP) and compared with conventional two-term Karatsuba multiplication (CTTKM) and existing state-of-the-art. Hardware implementations of both the proposed TTKM multiplication architectures are done using Virtex-7 FPGA device in Xilinx ISE platform. Compared with other state-of-the-art designs the performance of the proposed two-term Karatsuba multiplication (Method-I) based on ΔATP 1 is 11.108%, 98.686%, 15.561%, 72.012%, 95.122%. 6.009% and 24.601% better than Direct Multiplication (DM), Karat-suba Direct Multiplication (KDM), Karatsuba Comba Multiplication (KCM), Traditional Schoolbook Multiplication (Traditional SBM), SBM-I, SBM-II and CTTKM respectively for 512-bit inputs. Similarly the performance of the proposed two-term Karatsuba multiplication (Method-II) based on ΔATP 2 is 1.790%, 98.544%, 6.407%, 68.978%. 94.593% and 16.427% better than Direct Multiplication (DM), Karatsuba Direct Multiplication (KDM), Karatsuba Comba Multiplication (KCM), Traditional SBM, SBM-I, SBM-II, CTTKM respectively for 512-bit inputs. However comparing both the proposed methods of TTKM (Method-I and Method-II), it can be inferred that TTKM-I is 9.780% better than TTKM-II thus proving its area-time-product efficiency.

In this paper, we propose a novel encryption algorithm for medical images. Our algorithm is based around ad-vanced encryption standard (AES), however, its simple algebraic structure and the encryption process of every block may become a major concern in security. This is because the block-wise nature of implementation can pass the features of the actual image to visible in the encrypted image. We addressed this issue by introducing two new encryption layers i.e., Fractional Discrete Cosine Transform (FrDCT) and Arnold Transform. unlike the block-wise AES, these two layers are applied image-wide. together this cascade provides excellent security and key sensitivity while preserving the quality of the medical images. The proposed method has been compared with contemporary methods in both quantitative and qualitative manner which further validated our ideas. Even slight change in keys $(\Delta\approx\pm 0.001)$ causes a failure in decryption. This shows the robustness of the proposed algorithm.

In this work, we propose a Radial derivative Gaussian feature (RDGF) descriptor, a novel handcrafted feature descriptor for disguised thermal face recognition. The feature encoding has been done so that the performance is least affected by noise and works well over challenging datasets. We propose a cascaded framework that combines two modules, namely BoCNN and the RDGF descriptor. The cascading architecture estimates the performance of BoCNN before classification. It also uses a dynamic classifier selector in run time to choose between handcrafted features and the CNN framework to enhance the overall performance. We also propose a thermal face dataset with partial occlusion. We have compared the performance of the RDGF descriptor with state-of-the-art descriptors on the IIIT-Delhi disguised thermal face dataset and our proposed dataset. RDGF exhibits better performance compared to other state-of-the-art descriptors. Our proposed descriptor shows relative increment of 56.84%, 64.92%, 67.25%, 64.03%, 48.06%, and 7.28% on IIIT-D Occluded Thermal Dataset when compared with LBP, LDP, LBDP, LVP, LGHP, and HOG, respectively. A similar enhancement of accuracy has been observed on our proposed dataset as well. An exhaustive comparison based on the performance of the cascaded framework with state-of-the-art CNN models has also been done in a similar fashion.

Usually, the current generated by the photodiode proportional to the oxygenated blood in the photoplethysmography (PPG) and functional infrared spectroscopy (fNIRS) based recording systems is small as compared to the offset-current. The offset current is the combination of the dark current of the photodiode, the current due to ambient light, and the current due to the reflected light from fat and skull . The relatively large value of the offset current limits the amplification of the signal current and affects the overall performance of the PPG/fNIRS recording systems. In this paper, we present a mixed-signal auto-calibrated offset current compensation technique for PPG and fNIRS recording systems. The system auto-calibrates the offset current, compensates using a dual discrete loop technique, and amplifies the signal current. Thanks to the amplification, the system provides better sensitivity. A prototype of the system is built and tested for PPG signal recording. The prototype is developed for a 3.3 V single supply. The results show that the proposed system is able to effectively compensate for the offset current.

In this work, geometry optimization of mechanical truss using computer-aided finite element analysis is presented. The shape of the truss is a dominant factor in determining the capacity of load it can bear. At a given parameter space, our goal is to find the parameters of a hull that maximize the load-bearing capacity and also don't yield to the induced stress. We rely on finite element analysis, which is a computationally costly design analysis tool for design evaluation. For such expensive to-evaluate functions, we chose Bayesian optimization as our optimization framework which has empirically proven sample efficient than other simulation-based optimization methods. By utilizing Bayesian optimization algorithms, the truss design involves iteratively evaluating a set of candidate truss designs and updating a probabilistic model of the design space based on the results. The model is used to predict the performance of each candidate design, and the next candidate design is selected based on the prediction and an acquisition function that balances exploration and exploitation of the design space. Our result can be used as a baseline for future study on AI-based optimization in expensive engineering domains especially in finite element Analysis.

This paper proposes a new cryptographic algorithm based on fractional discrete cosine transform (FrDCT) and discrete orthonormal Stockwell (DOST) transform to secure multimedia data. Fractionality makes the cryptosystem more reliable and ensures the security of the data transmitted in the unsecured channel. The proposed method describes the advantage of using FrDCT instead of fractional Fourier transform (FrFT). At the same time, the energy compaction property of DOST guarantees lesser memory and storage requirements. The proposed method has been compared with state-of-the-methods to ensure practicality in different applications. The standard quantitative parameters such as information entropy (≈ 8), NPCR (≈ 99%), UACI (≈ 33%) etc., show its immunity against unauthenticated user(s).

In wake of COVID-19, the world has adapted to a new order. People have started wearing mask on their faces to prevent getting infected. The present face recognition models are no longer proving to be efficient in the current circumstances. This is because, most of the informative part of the face is covered by mask. The periocular recognition therefore holds the key to future of face recognition. However, the periocular region proves to be insufficiently enough to generate highly discriminative features. Also, most of the pre-COVID-19 algorithms fail to work in cases, where the number of training images available is very less. We propose a lightweight periocular recognition framework that uses thermo-visible features and ensemble subspace network classifier to improve upon the existing periocular recognition systems named as Masked Mobile Lightweight Thermo-visible Face Recognition (MmLwThV). The framework successfully improves the accuracy over a single visible modality by mitigating the effect of noise present in the thermo-visible features. The experiments on WHU-IIP dataset and an in-house collected dataset named, CVBL masked dataset, successfully validate the efficacy of our proposed framework. The MmLwFR framework is lightweight and can be easily deployed on mobile phones with a visible and an infrared camera.

In order to accurately predict the current health condition of rotating machinery tools, multiple sensors are mounted on rotating machines to capture diverse fault signals. These raw signals are then subjected to statistical feature extraction to predict the bearing's health condition. However, variations in the information of collected sensor signals occur due to sensor placement and structural/environmental noises that potentially lead to inconsistent prediction results. To address this problem, a novel method for extracting features from multisensor data is proposed. In the proposed method, first, spectrogram-based features have been extracted from different sensor data. Correlated features are then extraced from these features using the deep canonical correlation model. Further, feature pooling has been performed to get discriminative features. Finally, pooled feature vectors are used to train the convolutional neural network to recognize the health state of the machine. The PRONOSTIA degradation-bearing datasets are used to validate the proposed methodology. The datasets contain vibration data that has been collected across varying running speeds and loads. The proposed approach has also been tested under noisy and imbalanced conditions. We demonstrate that our method can effectively recognize the machine's fault conditions.

Abstract In this paper, a novel color image encryption scheme based on fractional discrete cosine transform (FrDCT) and discrete wavelet transform (DWT) with interplane arrangements in the discrete orthonormal Stockwell transform (DOST) domain is presented. Color images are encrypted using the keys of FrDCT and coefficients of interplane arrangement of DWT sub‐bands. These keys are applied on three independent planes, that is, R, G, and B. To ensure the correct decryption of the encrypted image, every key need to be clearly understood in the correct sequence and its exact values. FrDCT has the advantageous feature of storing real‐valued coefficients, and the energy of the image can be represented with very few coefficients. Therefore, transmitting and storing the encrypted image becomes easy. FrDCT may be considered a real‐valued cosine version of fractional Fourier transform (FrFT), which is used successfully for storage and transmission due to its property of compacting energy distribution. Furthermore, for more robustness, interplane arrangements are explored. The concept of interplane arrangements has not been substantially reported in the existing literature. The comparison of the technique with other similar encryption techniques available in the literature has been carried out by performing different experiments to examine the efficacy of the proposed method. It is found that the proposed method outperforms other existing techniques.

This paper investigates weak signal detection using approximated fractional integrator (AFI). An optimal spectrum sensing scheme based on the criterion of maximization of the deflection ratio coefficient (DR) has been proposed in the frequency domain. Due to the easy availability and less time complexity of fast Fourier transform (FFT) chips in the industry, signal detection is primarily done in the frequency domain. Approximated fractional integrator transforms the data to give the maximum DR because of its non-linear low-pass characteristic. Example and simulation results have been illustrated to observe the performance and properties of the proposed detector. A precise comparison has been discussed with state-of-the-art techniques.

Optical sensors are widely used in a variety of industrial, scientific and healthcare applications. The offset current due to ambient light affects the overall performance of such sensor system. This is even more critical in biomedical applications such as photoplethysmography (PPG) and functional-near infrared spectroscopy (fNIRS). The continuous offset cancellation technique introduce the delay and affects the shape morphing of such signals. In this paper, we present a mixed-signal based discrete offset cancellation technique for effective compensation of the offset current of the optical sensors for biomedical applications. The system is based on a feedback loop which calibrate the offset and compensate it at the start of recording. A prototype of the system is built and tested for PPG signal recording. The results shows that the proposed system is able to effectively compensate the offset current.

The ever-increasing mortality rate due to cancer has necessitated the development of new and effective techniques for selective destruction of cancerous cells. Among the various therapeutic methods, photo-thermal therapy has attracted considerable attention. The subject of photo-thermal therapy requires a detailed understanding of light-tissue interaction and the subsequent heat transfer process(es) through biological samples. The present review article focuses on the detailed review of literature on numerical modeling of the phenomena of light-tissue interaction and the resultant thermal response of laser-irradiated biological samples. While a comprehensive review of works of various researchers has been presented as a literature survey, discussion on various numerical models and the corresponding results has been primarily based on the authors' published works. Following the principles of photo-thermal therapy, the temperature rise of biological tissue depends on the light absorption, which makes it essential to precisely model the phenomenon of laser-tissue interaction. In this context, the transient radiative transfer equation is considered to be the most accurate. Being an integrodifferential equation, complete analytical solution of the radiative transfer equation (RTE) is challenging. Hence, a range of numerical models was developed to determine the light intensity distribution within the laser-irradiated biological samples. In this direction, the transient RTE has been solved using the discrete ordinates method (DOM). The solution of the RTE has then been coupled with various bio-heat transfer models. In order to quantify the influence of convective effects on tissue temperature distribution, the pulsatile nature of blood flow in large blood vessels has been considered. The solution of the RTE has been coupled with the energy equation, and Navier-Stokes equations have been solved for the velocity field. Results of the study revealed a strong influence of the pulsatile blood flow on the temperature distribution in the surrounding tissue region. An increase in temperature due to laser-irradiation is found to be less in the presence of blood flow as compared to that achieved without blood vessels. Limitations of conventional Fourier heat conduction models in accurately predicting the temperatures of laser-irradiated biological samples have been highlighted by various researchers. These limitations arise primarily due to the assumption of the infinite speed of thermal wave propagation. Such assumptions break down in biological samples since they are primarily composed of nonhomogeneous structures. These aspects have emphasized the importance of non-Fourier heat conduction models for photo-thermal applications. With this as the motivation, the solution of the transient RTE has been coupled with the generalized form of non-Fourier heat conduction models to predict the thermal response of the tissue phantoms. The non-Fourier numerical results have also been compared to the corresponding finite integral transform (FIT)-based analytical solutions. The relative influence of relaxation times associated with the temperature gradients (τT) and heat flux (τq) on the resultant thermal profiles has been studied and discussed. The work reported in this review article holds importance in optimizing the laser parameters for therapeutic applications so that the cell destruction is limited only up to the extent of abnormal/cancerous cells and minimum damage is incurred to the surrounding normal/healthy biological cells.

Thermal cameras can capture images even in low light conditions. However, humans cannot recognize human faces in thermal images. Translation of thermal images to visible domain is one solution to the problem of face recognition in thermal images. Most of the research works have proposed Generative Adversarial Networks (GANs) based solutions for thermal to visible image translation. However, GAN is a heavy network that consumes huge amount of resource for thermal to visible image translation. In this paper, we propose an encoder-decoder architecture for thermal to visible image translation of human faces. Since our proposed architecture is not based on GANs, it is lightweight. The proposed method works well for both disguised and non-disguised thermal facial images. Standard comparison parameters such as Peak Signal-to-noise Ratio (PSNR), Structural Similarity Index (SSIM), and Multiscale Structural Similarity Index (MS-SSIM) are used to evaluate the quality of the generated visible images with respect to the ground truth. It has been found that our proposed architecture outperforms the current state-of-the-art image translator architectures namely pix2pix, Cycle-GAN, modified thermal to visible GAN and Dual GAN by a considerable margin for both disguised as well as non-disguised dataset. (c) 2022 Elsevier B.V. All rights reserved.

For the quick and accurate dynamic response of a sensorless drive system, proper estimation of position and speed is indispensable. A conventional sliding mode observer (SMO) based sensorless control has several limitations like undesired frequency swing during phase jump, spurious noise speed reversal issue. This paper presents an improved SMO based sensorless control and finite control set model predictive control (FCS-MPC) of permanent magnet synchronous motor (PMSM) EV drive with regenerative braking. An improved phase locked loop (PLL) with a modified control function is utilized to obtain position and speed. It provides a smooth startup and avoids transient in frequency and singularities issues. To avoid cascading linear controllers, a robust FCS-MPC is employed in this paper. It provides switching pulses directly to the switches of voltage source inverter (VSI), which avoids the use of modulators. The proposed sensorless control method for an electric vehicle (EV) application is simulated using Simulink and the effectiveness and compatibility of this control algorithm is verified.

The active thermal non-destructive testing and evaluation technique plays a vital role in health monitoring of various solid materials. Present manuscript demonstrates the applicability of pulse compression favorable Digitized version of linear Frequency Modulated Thermal Wave Imaging (DFMTWI) approach to identify flaws having different geometrical shapes in a Glass Fibre Reinforced Polymer (GFRP) sample. A novel Thermal Image Correlation (TIC) data-processing approach is proposed to obtain the isothermal patterns from the reconstructed pulse compressed data through matched filter scheme to identify sub-surface anomalies. The detection capabilities of the presented approach are compared on various adopted data processing approaches.

Nowadays, securing biomedical images from several attacks have become an important issue. Biomedical images are highly sensitive, and it requires high security and copyright protection in telemedicine applications. Watermarking is the perfect choice for this purpose. Watermarking is a technique of hiding signature information in an original signal where the signature can be in the form of an image, random message, or a video. In this article, we have proposed a new, robust and highly secured technique of reversible watermarking in biomedical imaging that does not affect the image quality and readability. In this technique, the Region of Interest (ROI) and the Region of Not Interest (RONI) are separated from the original image. Watermark embedding is preferred in RONI. Watermark signal can be extracted with high quality from the watermarked image. This article discusses the pros and cons of current methods of watermarking in medical imaging. This method gives a high value of peak signal to noise ratio (PSNR) and the low value of root mean square error (RMSE). The technique does not alter any useful information of the original image and can be used against different categories of attack. It provides proof of ownership and full control over images.

We developed an architecture considering the problem of the presence of various objects like glasses, scarfs, masks, etc. on the face. These types of occluders greatly impact recognition accuracy. Occluders are simply treated as a source of noise or unwanted objects in an image. Occlusion affects the appearance of the image significantly. In this paper, a novel architecture for the occlusion removal of specific types like glasses is dealt with. A synthetic mask for this type of occlusion is created and facial landmarks are generated which are further used for image completion method is used to complete the image after removing the occluder. We have trained the model on celebA dataset, which is available publicly. The experimental results show that the proposed architecture for the occlusion removal effectively worked for the faces covered with glasses or scarfs.

This paper presents an efficient sensorless permanent magnet synchronous motor (PMSM) drive based on model reference adaptive system (MRAS) speed and position estimation for solar photovoltaic (PV)- battery power driven electric vehicle (EV). The aim of this work is to include a technique that encourages renewable energy applications. This work involves regenerative braking to increase the range of EV. The kinetic energy of the vehicle is recovered during braking conditions and utilized to charge the battery as well as to avoid deep discharge of the battery during the uphill driving condition. In order to reduce cost, weight, and increase robustness of the drive, MRAS based sensorless control is used for speed and position estimation. In this method, a current adaptable model is analyzed in which, PMSM itself acts as a reference model. A solar PV array integrated with a battery provides a seamless and continuing operation independent of environmental conditions. A Cuk DC-DC converter is utilized for maximum power point tracking (MPPT), which ensures continuous input and output currents. This system is efficient, cost-effective, and suits for practical implementation. A study of the PMSM control and the effectiveness of the drive system simulated using MATLAB simulink is presented in this paper.