We investigate the problem of infection diagnosis in hydrocephalic images. While traditional deep learning methods can be used for this problem - they require generous training data and produce class activations that are not meaningful. A novel 2D brain attention regularized network (BAR-Net) is proposed in our work, which encourages the network to focus inside the brain region for classification decisions. The newly proposed regularizer yields more interpretable class activations and experimental results confirm that the issue of over-fitting is mitigated. Extending the 2D BAR-Net, a hybrid 2-D/3-D CNN is proposed, that facilitates interactions between the 2D and 3D branches. Furthermore, a mutual attention regularization term is introduced in order to enable branches to exchange knowledge learned from the other by transferring attention maps for regularization.

The pandemic has precipitated a shift of value creation toward economic activity that is ignored in measures of economic growth such as GDP. As a consequence, economic policy is increasingly fumbling in the dark. Not least innovation policies and growth-oriented R & D now lack a relevant yardstick of success. This article provides an overview of the challenge and suggests better ways to measure the impact of innovations. Ultimately this concerns what kind of development nations aim for and how they assess success.

Semantic communications target to reliably convey the semantic meaning of messages. It is different from existing communication systems focusing on reliable bit transmission. To achieve the goal of semantic communications, we propose a signal shaping method by minimizing the semantic loss, which is measured by the pretrained bidirectional encoder representation from transformers (BERT) model. The signal set optimization problem is transformed to a vector optimization subject to a power constraint. We propose an efficient projected gradient descent method to solve the problem and prove its convergence. Simulation results show that the proposed method outperforms existing signal shaping methods in minimizing the semantic loss.

Internet-of-drones (IoD) systems require enhanced data transmission security and efficient interference management to accommodate the rapidly growing drone-based and rate-intensive applications. This paper develops a novel resource allocation scheme to jointly manage interference and enhance the physical layer security  of  cellular-connected IoD networks in the presence of a multi-band eavesdropping drone. Our envisioned cellular-connected IoD network has multiple full-duplex cellular base stations (CBSs), where each CBS reserves an orthogonal cellular  radio resource block (RRB) for the aerial communication links. To efficiently utilize the cellular RRBs, each CBS is connected to a cluster of data transmitting drones using uplink  non-orthogonal multiple access (NOMA) scheme. In addition, all the CBSs simultaneously transmit  artificial noise signals to weaken the eavesdropper links.  A joint optimization problem, considering the transmit power allocation and clustering of the legitimate drones, and the jamming power allocation of the CBSs, is formulated to maximize the worst-case average sum-secrecy-rate of the  network. The joint optimization problem  is decomposed into  drone-clustering and power allocation sub-problems to obtain an efficient solution. A multi-agent reinforcement-learning framework is devised to solve the drone-clustering sub-problem.  Meanwhile, the transmit and jamming power allocation sub-problem is solved by employing fractional programming, successive convex approximation, and alternating optimization techniques. By iteratively solving these two sub-problems, a convergent resource allocation algorithm, namely, security and interference management with reinforcement-learning and NOMA (SIREN), is proposed.  The superiority of SIREN over several benchmark schemes is verified via extensive simulations.

Short Summary: The paper aims to examine how the use of a variable illuminant can be used to affect colour perception and simulate deficient colour vision in people with typical colour perception. A 12 colour channel light source is used to have precise control over the illuminant’s spectral composition and the observed effects are contrasted against digital simulations of colour deficiencies. Paper Abstract: Normal human colour vision relies on 3 different cone photoreceptors (L, M, and S) in the retina and 2 opponent processes: red-green and blue-yellow. However, the exact hue perceived when observing an object is dependent upon three main factors, namely the genes encoding the opsins present in the observer’s eye, the spectral reflectivity of the object as well as the spectral composition of the incident light, also known as the illuminant. By controlling the composition of the illuminant, the amount of available light that can be reflected by an object can be controlled, which in turn affects the size of the observer’s colour gamut. By choosing illuminant spectra mostly/only containing colours that are most prominent to people affected by colour deficiencies, we can replicate the reduced colour gamut that they observe. Conversely, this technique can also be used to enhance certain colours and equalise the brightness of hues across the spectrum.

Spectrum sensing in cognitive radio (CR) paradigm can be broadly categorized as analytical based and data-driven approaches. The former is sensitive to model inaccuracies in evolving network environment, while the latter (machine learning (ML)/deep learning (DL) based approach) suffers from high computational cost. For devices with low computational abilities, such approaches could be rendered less useful. In this context, we propose a deep unfolding architecture namely the Primary User-Detection Network (PU-DetNet) that harvests the strength of both: analytical and data-driven approaches. In particular, a technique is described that reduces computation in terms of inference time and the number of floating point operations (FLOPs). It involves binding the loss function such that each layer of the proposed architecture possesses its own loss function whose aggregate is optimized during training. Compared to the state-of-the-art, experimental results demonstrate that at SNR= -10 dB, the probability of detection is significantly improved as compared to the long short term memory (LSTM) scheme (between 39% and 56%), convolutional neural network (CNN) scheme (between 45% and 84%), and artificial neural network (ANN) scheme (between 53% and 128%) over empirical, 5G new radio, DeepSig, satellite communications, and radar datasets. The accuracy of proposed scheme also outperforms other existing schemes in terms of the F1-score. Additionally, inference time reduces by 91.69%, 90.90%, and 93.15%, while FLOPs reduces by 62.50%, 56.25%, 64.70% w.r.t. LSTM, CNN and ANN schemes, respectively. Moreover, the proposed scheme also shows improvement in throughput by 56.39%, 51.23%, and 69.52% as compared to LSTM, CNN and ANN schemes respectively, at SNR = -6 dB.

Development of ultra-compact, low-to-medium precision analog-to-digital converters (ADCs) with unprecedented energy-efficiency is essential to meet the ever-increasing demand for data converters in advanced computing systems including neuromorphic accelerators based on emerging non-volatile memories. To this end, in this work, for the first time, we propose a feedforward neural network ADC based on a network of highly scalable, CMOS-compatible, and energy-efficient ferroelectric-FinFET (Fe-FinFET) synaptic elements. Our lower triangular neural network (LTNN) ADC design, implemented using 7-nm technology from ARM along with an experimentally calibrated compact model for Fe-FinFETs, consumes 5.44 μW of power, 1.03 μm2 of area while operating at a speed of 1.23 megasamples per second for 4-bit precision. The proposed neural network ADC may pave the way for realization of highly efficient neuromorphic processing engines and neuro-optimizers based on cross-point array of emerging non-volatile memories.

Despite the use of High-k gate stacks, poor channel electrostatics at short channel lengths has limited the applicability of conventional Bulk MOSFETs. With the use of Ferroelectric materials in the gate stack, enhanced transverse (gate) electric field has the potential to enable greater scalability, which needs to be further explored in the context of planar Bulk MOSFETs. In this paper, we firstly present the design of an optimized N-channel Partially Junction-less Bulk MOSFET with a view to suppressing various leakage mechanisms, while also taking channel quantization effects into account. Through detailed process simulations, the process steps required to achieve this device design are determined, thus providing a more practical appraisal of the feasibility of the device. We then identify the thickness of the Ferroelectric layer over the existing gate stack with a view to achieve steep sub-threshold slope (SS) while simultaneously ensuring hysteresis-free characteristics, including enhanced Analog as well as Digital performance of the device.

Abstract—Aiming to address the growth and the integration of renewable sources and electronic loads, different types of dc-dc topologies with specific features have been proposed, going from medium frequency isolated Modular Multilevel Converter to Switched Capacitor and Hybrid Switched Capacitor topologies. In this paper, a bidirectional Modular Multilevel Converter based Hybrid Switched Capacitor dc-dc converter is presented, featuring transformerless structure, automatic total arm voltages clamping/balancing, voltage static gain independent of the submodules quantity and efficiency with reduced duty cycle influence. The topology operates with a Quasi-Two-Level modulation, which provides passive components volume/weight reduction, a degree of freedom regarding the dv/dt applied on the magnetic windings and basis to perform the submodule capacitor voltages equalization within each arm. The topology operating principles are approached by an average value static analysis while an instant value static analysis is used to estimate the operating current spike worst case. A 1.17 kW, 350/200 V laboratory scale prototype using IGBTs is implemented to experimentally confirm the theoretical approaches.

Traditional convolutional neural network (CNN) methods rely on dense tensors, which makes them suboptimal for spatially sparse data. In this paper, we propose a CNN model based on sparse tensors for efficient processing of large and sparse medical images. In contrast to a dense CNN that takes the entire voxel grid as input, a sparse CNN processes only on the non-empty voxels, thus reducing the memory and computation overhead caused by the sparse input data. We evaluate our method on two clinically relevant skull reconstruction tasks: (1) given a defective skull, reconstruct the complete skull (i.e., skull shape completion), and (2) given a coarse skull, reconstruct a high-resolution skull with fine geometric details (shape super-resolution). Our method outperforms the state of the art in the skull reconstruction task quantitatively and qualitatively, while requiring substantially less memory for training and inference. We observed that, on the 3D skull data, the overall memory consumption of the sparse CNN grows approximately linearly during inference with respect to the image resolutions. During training, the memory usage remains clearly below increases in image resolution - an x8 increase in voxel number leads to less than x8 increase in memory requirements. Our study demonstrates the effectiveness of using a sparse CNN for skull reconstruction tasks, and our findings can be applied to other spatially sparse problems. We proof this by additional experimental results on other sparse medical datasets, like the aorta and the heart. Project page at https://github.com/Jianningli/SparseCNN

This work presents a formal review of smart contracts, including definitions, technical requirements, and potential power and energy-related use cases. This includes in-depth discussions covering cybersecurity, legality and interoperability goals that must be taken into consideration by potential end-users. The paper presents a first attempt towards the standardization of smart contracts (SCs) within the field of power and energy as a work in progress activity under the IEEE Standards Association (IEEE SA) P2418.5 Working Group. This work also proposes a holistic, language-agnostic reference model that is intended to accelerate the adoption of Distributed Ledger Technology (DLT) by industry stakeholders by providing standardized processes. Finally, the paper discusses key takeaways that must continue to be developed to increase SC usage within the energy industry.

This communication presents a corrected formulation, differing significantly from a traditional and widely employed one, for the precision of topographic measurements by means of space-based or airborne radar interferometry. The reasons for the differences are explained. The results are applicable to both downlooking topography and conventional geometries for interferometric synthetic aperture radar. Comparisons between the opposing formulations are exhibited for interferometer baseline tilt angles relative to the horizontal of five and sixty degrees. The discrepancies are especially wide at near-nadir viewing. The corrected formulation for measurement precision is presented in both its exact form and a convenient approximate expression, the two differing by less than one percent for the spaceborne interferometer illustrated.

This review explores mild cognitive impairment support, focusing on virtual reality (VR) applications coupled with electrodermal activity (EDA). A systematic review and meta-analysis using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement were performed from an original inquiry of 698 articles from 11 digital libraries. An independent review of selected studies assessed the risk of bias and obtained data. The meta-analysis included 44 studies, with a total number of 9099 participants in all the papers reviewed were split into (5200) cognitive or (2304) non-cognitive impairments. The proportion of cognitive versus non-cognitive had a first risk difference (RD = 0.32; p-value = 0.0001). The comparison of female to male participants showed (OR = 6.10; p-value = 0.00001). Anxiety, depression, and stress conditions had the effect of (SMD = -0.39; p-value = 0.09). In the analysis, the general lack of research restricts virtual reality and electrodermal activity to sustain mild cognitive impairments. Forty-five excluded papers applied data from pre-2017. Two studies with 46 participants had a result of (SMD = 0.60; p-value = 0.31). Identifying VR and EDA-based studies investigating training and rehabilitation to support MCI symptoms in individuals.

The most known of public key cryptosystem was introduced in 1978 by Rivest, Shamir and Adleman [19] and now called the RSA public key cryptosystem in their honor. Later, a few authors gave a simply extension of RSA over algebraic numbers field( see [20]- [22]), but they require that the ring of algebraic integers is Euclidean ring, this requirement is much more stronger than the class number one condition. In this paper, we introduce a high dimensional form of RSA by making use of the ring of algebraic integers of an algebraic number field and the lattice theory. We give an attainable algorithm (see Algorithm I below) of which is significant both from the theoretical and practical point of view. Our main purpose in this paper is to show that the high dimensional RSA is a lattice based on public key cryptosystem indeed, of which would be considered as a new number in the family of post-quantum cryptography(see [17] and [18]). On the other hand, we give a matrix expression for any algebraic number fields (see Theorem 2.7 below), which is a new result even in the sense of classical algebraic number theory.

We experimentally demonstrate bandwidth-tunable RF photonic Hilbert transformer based on an integrated Kerr micro-comb source. The micro-comb is generated by an integrated micro-ring resonator with a free spectral range of 48.9 GHz, yielding 75 micro-comb lines in the telecom C-band. By programming and shaping the generated comb lines according to calculated tap weights, we demonstrate high-speed Hilbert transform functions with tunable bandwidths ranging from 1.2 GHz to 15.3 GHz, switchable center frequencies from baseband to 9.5 GHz, and arbitrary fractional orders. The experimental results show good agreement with theory and confirm the effectiveness of our approach.

Integrated photonic devices operating via optical nonlinearities offer a powerful solution for all-optical information processing, yielding processing speeds that are well beyond that of electronic processing as well as providing the added benefits of compact footprint, high stability, high scalability, and small power consumption. The increasing demand for high-performance nonlinear integrated photonic devices has facilitated the hybrid integration of novel materials to address the limitations of existing integrated photonic platforms, such as strong nonlinear optical absorption or an inadequate optical nonlinearity. Recently, graphene oxide (GO), with its large optical nonlinearity, high flexibility in altering its properties, and facile fabrication processes, has attracted significant attention, enabling many hybrid nonlinear integrated photonic devices with improved performance and novel capabilities. This paper reviews the applications of GO to nonlinear integrated photonics. First, an overview of GO’s optical properties and the fabrication technologies needed for its on-chip integration is provided. Next, the state-of-the-art GO nonlinear integrated photonic devices are reviewed, together with comparisons of the nonlinear optical performance of different integrated platforms incorporating GO. Finally, the challenges and perspectives of this field are discussed.

Shortages in therapists, along with the COVID-19 pandemic, have created a gap in care for people with cerebral palsy, stroke, and many other conditions requiring rehabilitation. Socially assistive robots and telehealth are two tools which could help fill this gap and may be able to work together, in socially assistive robot-augmented telerehabilitation, to create meaningful telerehab encounters. We present results from a survey completed between Jun 31, 2020, and Dec 31, 2020, asking 379 therapists in the United States for their opinions on features which may be useful for a telerehabilitation platform and for their general feelings towards robots. We explore their answers and reduce the response space by identifying six factors underlying the answers. We show that therapists want high quality video, control over robots, and more regulations. We also find that they are positive about the features which could be used in socially assistive robot-augmented telerehabilitation.

Early detection of an inter-turn short circuit fault (ISCF) can reduce repair costs and downtime of an electrical machine. In an induction machine (IM) driven by an inverter with a model predictive control (MPC) algorithm, the controller outputs are influenced by a fault due to the fault-controller interaction. Based on this observation, this study developed neural network models using inverter switching statistics to detect the ISCF of an IM. The method was non-invasive, and it did not require any additional sensors. In the fault detection task, an area under receiver operating characteristics curve value of 0.9950 (95% Confidence IntervaI: 0.9949 - 0.9951) was obtained. At the rated operating conditions, the neural network model detected and located an ISCF of 2-turns (out of 104 turns per phase) under 0.1 seconds, a speedup of more than two times compared to the thresholding-based method. Moreover, we published the switching vector data collected at various load torque and shaft speed values for healthy and faulty states of the IM, becoming the first publicly available ISCF detection dataset. Together with the dataset, we provided performance baselines for three main neural network architectures, namely, multi-layer perceptron, convolutional neural network, and recurrent neural network.