Measurements of cardiac function such as left ventricular ejection fraction and myocardial strain are typically based on 2D ultrasound imaging. The reliability of these measurements strongly depends on the correct pose of the transducer such that the 2D imaging plane properly aligns with the heart for standard measurement views, and is thus dependent on the operator’s skills. In this work, we propose a deep learning-based tool that provides real-time feedback on how to move the transducer to obtain the required views. We believe this method can aid less-experienced users to acquire recordings of better quality for measurements and diagnosis, and to improve standardization of images for more experienced users. Training data was generated by slicing 3D ultrasound volumes, which permits to simulate movements of a transducer and 2D imaging plane. Each slice was labelled with an anatomical reference obtained through a semi-automatic annotation procedure, which allowed us to generate substantial amounts of training data. The method was validated and tested on 2D images from several datasets representative of a prospective clinical setting. We proposed a new metric to score the correctness of the transducer movement feedback according to several given criteria, and achieved a success rate of 75% for all models and 95% for the rotational movement. A real-time prototype application was developed based on data streaming from a clinical ultrasound system, which demonstrated the ability of the method to robustly predict the apical rotation and tilt of the 2D ultrasound image plane relative to the heart.

A document by Dozie Ubosi . Click on the document to view its contents.

Amputees face significant challenges in performing daily activities after losing an upper limb. Replacing a lost limb with a device with capabilities comparable to a human hand is a keen area of interest for prosthetics researchers. Despite advances in the field of prosthetics, the number of amputees who use these devices remains low. This study presents an open-source, low-cost, and anthropomorphic prosthetic hand solution for people with upper-limb amputations that outperforms currently available devices in terms of usability and biomimetic performance. The work aims to improve the acceptance of these devices among amputees by reducing the cost and improving the performance of these devices. We briefly discussed how the adaptive force magnification technique could be applied to a prosthetic device to improve biomimetic performance characteristics such as gripping force and velocity. A non-backdriveable modular under-actuated mechanism is designed that synergistically moves the fingers to adapt according to the shape of the object. The developed device costs one order less than the cheapest available prosthetics in the market and is customizable according to the user. The experiments were conducted to evaluate the biomimetic performance characteristics and the grasping effectiveness of the device in performing activities of daily life.

The inverse source problem of near-field antenna measurements is often formulated in terms of unknown equivalent electric and magnetic surface current densities. Exploiting just the constraints for the radiated field exterior to a closed Huygens surface results in ambiguous equivalent sources due to the excitation of non-radiating currents during the solution process. Therefore, enforcing the zero-field or Love side condition for the suppression of non-radiating currents based on a left-hand side preconditioner in the form of a Calderón projector is proposed. Transformation results based on realistic measurements demonstrate the effectiveness of this approach.

Intelligent and autonomous networks require precise and fast mechanisms that ensure error-free and efficient operation. Modern solutions are increasingly based on  artificial intelligence, in particular on machine learning, to reliably process huge amounts of data. Therefore, high-quality datasets are essential to train machine learning models. Unfortunately, the problem of assessing the quality of datasets is very challenging and often overlooked. This paper proposes a method for assessing the dataset quality in the context of binary classification. It is based on permutation testing and examines the strength of the relationship between observations and labels. Experiments carried out on simulated and real network datasets show that the method is sensitive to detect errors/mislabels in the labeled dataset. We also present theoretical considerations justifying our results.

Our proposed system, Audibly, is a desktop app which records the input using microphone. The system then converts the input audio into text and processed.After processing text, the text will be split to an individual word. Each word will be searched in the ASL dictionary and its corresponding videoclip will be merged to output video. If not found, the word will be stemmed to its root word and searched again else videoclip of each alphabet will be used. ASL dictionary contains videoclips of more than 800 words and alphabets. Output video will be formed by merging sub videos in mp4 format. This video will be displayed in output page and users can if needed.

The accuracy problem of the classically Rao-Wilton-Glisson (RWG) discretized magnetic field and combined field integral equations originates from the strong singularity of the identity operator. With a novel basis transformation scheme, i.e. two subsequent weak-form rotations of the RWG basis functions, the accuracy of the identity operator discretization is improved. The presented discretization scheme is immediately compatible with fast algorithms and is combined with the multilevel fast multipole algorithm. The good accuracy and the good conditioning behavior of the formulation are verified in several numerical simulations, also including electrically large problems.

In this paper, we introduce new distributed diffusion algorithms to track a sequence of hidden random matrices that evolve on the special orthogonal group. The algorithms are based on the Adapt-then-Combine and the Random Exchange methods, and diffuse Gaussian approximations of posterior densities computed in the Lie algebra of the special orthogonal group. Simulation results show that, in a scenario with strongly nonlinear observation functions, the proposed algorithms perform similarly to the centralized particle filter estimator and can outperform competing Extended Kalman Filters.

The electric field integral equation (EFIE) for perfect electrically conducting objects is augmented by a weak form combined source side condition. Thereby, the solution of the resulting combined source integral equation (CSIE) is unique and the interior resonance problem is circumvented. The discretization of electric and magnetic surface current densities and the testing of the electric field is solely performed with Rao-Wilton-Glission functions. Simulation results of scattering by a sphere demonstrate the good iterative solver convergence and excellent accuracy, as compared to EFIE as well as magnetic and combined field integral equations.

The accuracy of near-field antenna measurements combined with near-field far-field transformations (NFFFT) is deteriorated if objects are located in close vicinity to the antenna under test (AUT). We compare two equivalent sources based modeling techniques to suppress the influence of structures in the vicinity of the AUT. Simulation results for perfect electrically conducting objects showcase the capabilities and limitations of both approaches. The presented methods can easily extend existing NFFFTs in order to remove the influence of mounting structures or parts of the positioning system in anechoic near-field measurement facilities.

In this paper, a novel reinforcement learning (RL) approach, continuous dynamic policy programming (CDPP) is proposed to tackle the issues of both learning stability and sample efficiency in the current RL methods with continuous actions. The proposed method naturally extends the relative entropy regularization from the value function-based framework to the actor-critic (AC) framework of deep deterministic policy gradient (DDPG) to stabilize the learning process in continuous action space. It tackles the intractable softmax operation over continuous actions in the critic by Monte Carlo estimation and explores the practical advantages of the Mellowmax operator. A Boltzmann sampling policy is proposed to guide the exploration of actor following the relative entropy regularized critic. Evaluated by several benchmark tasks, the proposed method clearly illustrates the positive impact of the relative entropy regularization including efficient exploration behavior and stable policy update in RL with continuous action space and successfully outperforms the related baseline approach in both sample efficiency and learning stability.

We proposed a universal clustering algorithm by constructing an adaptive density gradient. This algorithm showed no data preference, and performs well on data with arbitrary density, overlap and shape. In comparative experiments, it outperformed other state-of-the-art algorithms on all types of synthetic and real data.

Starting from the monopolar representation of the Rao-Wilton-Glisson (RWG) functions, an approximate inverse (AI) of the RWG Gram matrix is constructed, where only 3x3 matrices for all individual triangles have to be inverted. The inversion of the Gram matrix is, however, only approximate, since it results as the inverse of a product of singular matrices, which is not easily obtained. The AI is shown to be an effective preconditioner for the RWG Gram matrix and is in particular useful for ill-conditioned Gram matrics on distorted meshes.

Optical full-field recovery makes it possible to compensate fiber impairments such as chromatic dispersion and polarization mode dispersion (PMD) in the digital signal processing. Coherent detection dominates over the long-haul transmission system due to its high spectral efficiency and tolerance against optical impairments. For cost-sensitive short-reach optical networks, some advanced single-polarization (SP) optical field recovery schemes are recently proposed to avoid chromatic dispersion-induced power fading effect, and improve the spectral efficiency for larger potential capacity. Polarization division multiplexing (PDM) can further double both the spectral efficiency and the system capacity of these SP carrier-assisted direct detection (DD) schemes. However, the so-called polarization fading phenomenon induced by random polarization rotation is a fundamental obstacle to exploit the polarization diversity for carrier-assisted DD systems. In this paper, we propose a receiver of Jones-space field recovery (JSFR) to realize polarization diversity with SP carrier-assisted DD schemes in the Jones space. There different receiver structures of proposed JSFR and the simplification for JSFR are explored theoretically. The proposed JSFR pushes the SP DD schemes towards PDM without no extra optical signal-to-noise ratio (OSNR) penalty. In addition, the JSFR shows good tolerance to PMD since the optical field recovery is conducted before polarization recovery. In the concept-of-proof experiment, we demonstrate 448-Gb/s reception over 80-km single-mode fiber using the proposed JSFR based on 2×2 couplers. Qualitatively, we compare the optical field recovery in the Jones space and Stokes space from the perspective of the modulation dimension.

A non-reciprocal self-mixing receive antenna array is presented. First, the principle of the self-mixing concept is explained and it is found that self-mixing receivers achieve a large gain over a wide angular range, which is in contrast to reciprocity. In the second part, the design of a 4×2 antenna array operating from 34 GHz to 39 GHz including the receive amplifier and mixer is introduced. Measurements in an anechoic chamber verify the simulation results, showing promising results for future applications of self-mixing arrays with a wide angular receiving range.

Abstract: Tensor tracking which is referred to as online (adaptive) decomposition of streaming tensors has recently gained much attention in the signal processing community due to the fact that many modern applications generate a huge number of multidimensional data streams over time. In this paper, we propose an effective tensor tracking method via the tensor-train format for decomposing high-order incomplete streaming tensors. On the arrival of new data, the proposed algorithm minimizes a weighted least-squares objective function accounting for both missing values and time-variation constraints on the underlying tensor-train cores, thanks to the recursive least-squares filtering technique and the block coordinate descent framework. Our algorithm is fully capable of tensor tracking from noisy, incomplete, and high-dimensional observations in both static and time-varying environments. Its tracking ability is validated with several experiments on both synthetic and real data. Code: https://github.com/thanhtbt/ATT-miss/ Comment: Accepted by Signal Processing Journal Â

Amid the recent advances in robotics and machine learning, unmanned aerial vehicles (UAVs) have shown evident proliferation across various applications. Consequently, the involvement of UAVs in populated environments has progressively become inevitable, putting forward stringent safety and security measures. In this work, we develop a deep reinforcement learning-based UAV-navigation approach that blends decision making with behavioral intelligence. In particular, a reinforcement learning (RL) agent is trained to instruct the UAV on how to accomplish a goal-oriented task, while assuring the safety of the UAV and its surroundings. Upon arriving at the goal position, the RL agent slows the UAV down preparing it for landing. The safety of the UAV and the environment are attained through a robust collision avoidance capability, embedded into the RL-based navigation system and considers both static and dynamic obstacles in the environment. Training is exclusively carried out in simulation, where a high fidelity UAV controller model is used to perform the simulated maneuver. The proposed approach was tested in simulation and then shown to directly transfer to reality without explicit sim2real gap transfer techniques. Experimental results demonstrated the agent’s capability to achieve the navigation task with 90% success rate.

A low-order discretization scheme for the magnetic field integral equation (MFIE) with considerably improved accuracy is discussed and analyzed. The formulation is based on the classical form of the MFIE discretized with Rao-Wilton-Glisson functions and introduces weak-form basis transformations, which have recently been employed with great success within the context of the so-called combined source integral equation. Only the identity operator discretization, which is known to be a major error source, is modified in order to obtain highly accurate solutions. Numerical results are given to corroborate the effectiveness of the new scheme.