Deep Learning Scheme Research Articles

Multiagent reinforcement learning for autonomous driving in traffic zones with unsignalized intersections

In this work we present a multiagent deep reinforcement learning approach for autonomous driving vehicles that is able to operate in traffic networks with unsignalized intersections. The key aspects of the proposed study are the introduction of route-agents as the main building block of the system, as well as a collision term that allows the cooperation among vehicles and the construction of an efficient reward function. These have the advantage of establishing an enhanced collaborative multiagent deep reinforcement learning scheme that manages to control multiple vehicles and navigate them safely and efficiently-economically to their destination. In addition, it provides the beneficial flexibility to lay down a platform for transfer learning and reusing knowledge from the agents’ policies in handling unknown traffic scenarios. We provide several experimental results in simulated road traffic networks of variable complexity and diverse characteristics using the SUMO environment that empirically illustrate the efficiency of the proposed multiagent framework.

Automatic classification of histology images into normal/cancer class with pre-trained CNN

Deep-Learning-Scheme (DLS) based medical data assessment has been widely employed in recent years due to its improved accuracy. Our goal is to study the performance of the pre-trained DLS on RGB-scale breast-histology images. The implemented idea holds these phases; (i) Data collection, pre-processing and resizing, (ii) Training the DLS with chosen test-pictures, (iii) Testing and validating the performance of the DLS with 5-fold cross-validation. This investigation considered the breast-histology pictures for the study and binary classification is employed to achieve Normal/Cancer class grouping of images. The proposed work compared the classification performance of AlexNet, VGG16 and VGG19.The experimental outcome of this study authenticates that the AlexNet with the Random-Forest (RF) classifier helps to get a higher classification accuracy (>87%) compared to VGG16 and VGG19.

Deep Reinforcement Learning With Graph Representation for Vehicle Repositioning

On-demand ride services, e.g., taxi companies, ridesourcing/transportation network companies (TNCs), play a vital role in nowadays transportation modes. The relocation/repositioning of idle vehicles is an important operational problem for service providers. With the recent success of deep learning, a large and growing body of literature emerged on learning-based transportation problems, including the relocation problem. In this work, we study the idle vehicle relocation problem to achieve better drivers’ profits and customers’ satisfaction. In detail, we formulate the learning problem on real-world road networks while recent repositioning works under a deep reinforcement learning scheme formulate the environment as grids. Meanwhile, a graph formulation enables learning with Graph Neural Network (GNN), which shows promising results in recent transportation studies. We empirically show that preserving the road network structure and learning the data representation with GNN contribute to improving the repositioning decision policy. Our simulation experiments are conducted on two real-world datasets, in New York City and Austin, respectively. The results show our formulation’s advantages in terms of reductions of order rejection rates and passenger waiting time, and increasing of vehicle occupancy rates and driver revenues.

An accurate Alzheimer's disease detection using a developed convolutional neural network model

Alzheimer's disease indicates one of the highest difficult to heal diseases, and it is acutely affecting the elderly normal lives and their households. Early, effective, and accurate detection represents an important blueprint for minimizing Alzheimer's progression risk. The modalities of brain imaging can assist in identifying the abnormalities associated with Alzheimer's disease. This research presents a developed deep learning scheme, which is designed and implemented to classify the brain images into multiclass, namely very mild, moderate, mild, and non-demented. The proposed convolutional neural network (CNN) based detection model attained a high performance with an accuracy of 99.92%, considerably enhancing the results achieved via the pre-trained 16 layers in the visual geometric group (VGG16) model and the other related learning models. Consequently, this developed model can assist medical personnel by providing a facilitating tool to identify Alzheimer's disease stage and establishing a suitable medical treatment platform.

DMRF-UNet: A Two-Stage Deep Learning Scheme for GPR Data Inversion Under Heterogeneous Soil Conditions

Traditional ground-penetrating radar (GPR) data inversion leverages iterative algorithms that suffer from high computation costs and low accuracy when applied to complex subsurface scenarios. Existing deep learning-based methods focus on the ideal homogeneous subsurface environments and ignore the interference due to clutters and noise in real-world heterogeneous environments. To address these issues, a two-stage deep neural network (DNN), called DMRF-UNet, is proposed to reconstruct the permittivity distributions of subsurface objects from GPR B-scans under heterogeneous soil conditions. In the first stage, a U-shape DNN with first multi-receptive-field convolution (MRF-UNet1) is built to remove the clutters due to inhomogeneity of the heterogeneous soil. Then, the denoised B-scan from MRF-UNet1 is combined with the noisy B-scan to be inputted to the DNN in the second multi-receptive-field convolution (MRF-UNet2). MRF-UNet2 learns the inverse mapping relationship and reconstructs the permittivity distribution of subsurface objects. To avoid information loss, an end-to-end training method combining the loss functions of two stages is introduced. A wide range of subsurface heterogeneous scenarios and B-scans are generated to evaluate the inversion performance. The test results in the numerical experiment and the real measurement show that the proposed network reconstructs the permittivities, shapes, sizes, and locations of subsurface objects with high accuracy. The comparison with existing methods demonstrates the superiority of the proposed methodology for the inversion under heterogeneous soil conditions.

Complex‐Amplitude Metasurface Design Assisted by Deep Learning

AbstractMetasurfaces can enable powerful manipulations for electromagnetic waves, thus many exotic functionalities have been realized. However, constrained by the complexity of metasurface‐modulated complex amplitudes (amplitude and phase), the design of complex‐amplitude metasurfaces sets a high threshold for researchers because of the requirement of plenty of specialized knowledge. In this paper, a deep learning scheme that uses a forward surrogate network to assist complex‐amplitude metasurface design is proposed. The model is simple to construct and easy to converge in training. Accordingly, two complex‐amplitude multiplexing devices, which can simultaneously display a nanoprinting image at the device surface and one/two holographic images in the far field, are designed with the proposed network using the cross‐shaped meta‐atom. The results show that the demonstrated scheme can be used to design the complex‐amplitude metasurface devices easily and effectively once training of the network is completed, and the single‐layer smooth structure holds the advantage for the fabrication. The proposed method here is promising to realize designs of more complex‐amplitude meta‐devices, multi‐polarization, and multi‐wavelength multiplexing meta‐devices.

Si Microring Resonator Crossbar Array for On-Chip Inference and Training of the Optical Neural Network

Deep learning has been a powerful tool in various fields. Faced with the limits of the current electronics platform, optical neural networks (ONNs) based on Si programmable photonic integrated circuits (PICs) have attracted considerable attention as a novel deep-learning scheme with optical-domain matrix-vector multiplication (MVM). However, most of the proposed Si programmable PICs for ONNs have several drawbacks, such as low scalability, high power consumption, and lack of frameworks for training. To address these issues, we have proposed a microring resonator (MRR) crossbar array as a Si programmable PIC for an ONN. In this Article, we present a prototype of a fully integrated 4 × 4 MRR crossbar array for MVM and demonstrate a simple image classification task using this chip. Moreover, we propose on-chip backpropagation using the transpose matrix operation of the MRR crossbar array, enabling the on-chip training of the ONN. The proposed ONN scheme can establish a scalable, power-efficient deep-learning accelerator for both inference and training tasks.

A Deep Learning Scheme for Detecting Atrial Fibrillation Based on Fusion of Raw and Discrete Wavelet Transformed ECG Features.

Atrial fibrillation is the most common sustained cardiac arrhythmia and the electrocardiogram (ECG) is a powerful non-invasive tool for its clinical diagnosis. Automatic AF detection remains a very challenging task due to the high inter-patient variability of ECGs. In this paper, an automatic AF detection scheme is proposed based on a deep learning network that utilizes both raw ECG signal and its discrete wavelet transform (DWT) version. In order to utilize the time-frequency characteristics of the ECG signal, first level DWT is applied and both high and low frequency components are then utilized in the 1D CNN network in parallel. If only the transformed data are utilized in the network, original variations in the data may not be explored, which also contains useful information to identify the abnormalities. A multi-phase training scheme is proposed which facilitates parallel optimization for efficient gradient propagation. In the proposed network, features are directly extracted from raw ECG and DWT coefficients, followed by 2 fully connected layers to process features furthermore and to detect arrhythmia in the recordings. Classification performance of the proposed method is tested on PhysioNet-2017 dataset and it offers superior performance in detecting AF from normal, alternating and noisy cases in comparison to some state-of-the-art methods.

ICOS: A Deep Reinforcement Learning Scheme for Wireless-Charged MEC Networks

Computation offloading is an effective method in mobile edge computing (MEC) to relieve user equipment (UE) from the limited computation resource and battery capacity. Meanwhile, simultaneous wireless information and power transmission (SWIPT) can be applied to MEC to extend the operating time of the equipment. However, in multi-user network environment, diverse computation task requirements and changeable network channel states make it challenging to obtain offloading strategy timely and accurately. To address the issue, we propose an intelligent computation offloading scheme (iCOS) based on enhanced priority deep deterministic policy gradient (EPDDPG) algorithm to minimize the energy consumption of all the UEs by jointly optimizing the offloading decision, the central processing unit (CPU) frequency and the power split ratio in a dynamic SWIPT-MEC network. In particular, we improve the traditional fully-connected network structure to obtain both discrete and continuous action outputs, and accelerate neural network parameter updates by using prioritized experience tuples. Furthermore, we use dynamic voltage and frequency scaling (DVFS) technology to dynamically adjust the CPU frequency of local computing, and employ SWIPT technology to balance the charging and communication according to the obtained strategy. Simulation results show that the algorithm proposed in this paper can effectively reduce the energy cost of UEs, and complete more computation tasks within the delay limit.

A novel deep reinforcement learning scheme for task scheduling in cloud computing

A novel deep reinforcement learning scheme for task scheduling in cloud computing

Simultaneously optimizing sampling pattern for joint acceleration of multi-contrast MRI using model-based deep learning.

Acceleration of MR imaging (MRI) is a popular research area, and usage of deep learning for acceleration has become highly widespread in the MR community. Joint acceleration of multiple-acquisition MRI was proven to be effective over a single-acquisition approach. Also, optimization in the sampling pattern demonstrated its advantage over conventional undersampling pattern. However, optimizing the sampling patterns for joint acceleration of multiple-acquisition MRI has not been investigated well. To develop a model-based deep learning scheme to optimize sampling patterns for a joint acceleration of multi-contrast MRI. The proposed scheme combines sampling pattern optimization and multi-contrast MRI reconstruction. It was extended from the physics-guided method of the joint model-based deep learning (J-MoDL) scheme to optimize the separate sampling pattern for each of multiple contrasts simultaneously for their joint reconstruction. Tests were performed with three contrasts of T2-weighted, FLAIR, and T1-weighted images. The proposed multi-contrast method was compared to (i) single-contrast method with sampling optimization (baseline J-MoDL), (ii) multi-contrast method without sampling optimization, and (iii) multi-contrast method with single common sampling optimization for all contrasts. The optimized sampling patterns were analyzed for sampling location overlap across contrasts. The scheme was also tested in a data-driven scenario, where the inversion between input and label was learned from the under-sampled data directly and tested on knee datasets for generalization test. The proposed scheme demonstrated a quantitative and qualitative advantage over the single-contrast scheme with sampling pattern optimization and the multi-contrast scheme without sampling pattern optimization. Optimizing the separate sampling pattern for each of the multi-contrasts was superior to optimizing only one common sampling pattern for all contrasts. The proposed scheme showed less overlap in sampling locations than the single-contrast scheme. The main hypothesis was also held in the data-driven situation as well. The brain-trained model worked well on the knee images, demonstrating its generalizability. Our study introduced an effective scheme that combines the sampling optimization and the multi-contrast acceleration. The seamless combination resulted in superior performance over the other existing methods.

CardioNet: Automatic Semantic Segmentation to Calculate the Cardiothoracic Ratio for Cardiomegaly and Other Chest Diseases.

Semantic segmentation for diagnosing chest-related diseases like cardiomegaly, emphysema, pleural effusions, and pneumothorax is a critical yet understudied tool for identifying the chest anatomy. A dangerous disease among these is cardiomegaly, in which sudden death is a high risk. An expert medical practitioner can diagnose cardiomegaly early using a chest radiograph (CXR). Cardiomegaly is a heart enlargement disease that can be analyzed by calculating the transverse cardiac diameter (TCD) and the cardiothoracic ratio (CTR). However, the manual estimation of CTR and other chest-related diseases requires much time from medical experts. Based on their anatomical semantics, artificial intelligence estimates cardiomegaly and related diseases by segmenting CXRs. Unfortunately, due to poor-quality images and variations in intensity, the automatic segmentation of the lungs and heart with CXRs is challenging. Deep learning-based methods are being used to identify the chest anatomy segmentation, but most of them only consider the lung segmentation, requiring a great deal of training. This work is based on a multiclass concatenation-based automatic semantic segmentation network, CardioNet, that was explicitly designed to perform fine segmentation using fewer parameters than a conventional deep learning scheme. Furthermore, the semantic segmentation of other chest-related diseases is diagnosed using CardioNet. CardioNet is evaluated using the JSRT dataset (Japanese Society of Radiological Technology). The JSRT dataset is publicly available and contains multiclass segmentation of the heart, lungs, and clavicle bones. In addition, our study examined lung segmentation using another publicly available dataset, Montgomery County (MC). The experimental results of the proposed CardioNet model achieved acceptable accuracy and competitive results across all datasets.

A 3-miRNA Signature Enables Risk Stratification in Glioblastoma Multiforme Patients with Different Clinical Outcomes.

Malignant gliomas constitute a complex disease phenotype that demands optimum decision-making as they are highly heterogeneous. Such inter-individual variability also renders optimum patient stratification extremely difficult. microRNA (hsa-miR-20a, hsa-miR-21, hsa-miR-21) expression levels were determined by RT-qPCR, upon FFPE tissue sample collection of glioblastoma multiforme patients (n = 37). In silico validation was then performed through discriminant analysis. Immunohistochemistry images from biopsy material were utilized by a hybrid deep learning system to further cross validate the distinctive capability of patient risk groups. Our standard-of-care treated patient cohort demonstrates no age- or sex- dependence. The expression values of the 3-miRNA signature between the low- (OS > 12 months) and high-risk (OS < 12 months) groups yield a p-value of <0.0001, enabling risk stratification. Risk stratification is validated by a. our random forest model that efficiently classifies (AUC = 97%) patients into two risk groups (low- vs. high-risk) by learning their 3-miRNA expression values, and b. our deep learning scheme, which recognizes those patterns that differentiate the images in question. Molecular-clinical correlations were drawn to classify low- (OS > 12 months) vs. high-risk (OS < 12 months) glioblastoma multiforme patients. Our 3-microRNA signature (hsa-miR-20a, hsa-miR-21, hsa-miR-10a) may further empower glioblastoma multiforme prognostic evaluation in clinical practice and enrich drug repurposing pipelines.

Deep Learning for Joint Pilot Design and Channel Estimation in MIMO-OFDM Systems.

In MIMO-OFDM systems, pilot design and estimation algorithm jointly determine the reliability and effectiveness of pilot-based channel estimation methods. In order to improve the channel estimation accuracy with less pilot overhead, a deep learning scheme for joint pilot design and channel estimation is proposed. This new hybrid network structure is named CAGAN, which is composed of a concrete autoencoder (concrete AE) and a conditional generative adversarial network (cGAN). We first use concrete AE to find and select the most informative position in the time-frequency grid to achieve pilot optimization design and then input the optimized pilots to cGAN to complete channel estimation. Simulation experiments show that the CAGAN scheme outperforms the traditional LS and MMSE estimation methods with fewer pilots, and has good robustness to environmental noise.

A Deep Learning Approach for Quantifying Vocal Fold Dynamics During Connected Speech Using Laryngeal High-Speed Videoendoscopy.

Voice disorders are best assessed by examining vocal fold dynamics in connected speech. This can be achieved using flexible laryngeal high-speed videoendoscopy (HSV), which enables us to study vocal fold mechanics with high temporal details. Analysis of vocal fold vibration using HSV requires accurate segmentation of the vocal fold edges. This article presents an automated deep-learning scheme to segment the glottal area in HSV from which the glottal edges are derived during connected speech. Using a custom-built HSV system, data were obtained from a vocally healthy participant reciting the "Rainbow Passage." A deep neural network was designed for glottal area segmentation in the HSV data. A recently introduced hybrid approach by the authors was utilized as an automated labeling tool to train the network on a set of HSV frames, where the glottis region was automatically annotated during vocal fold vibrations. The network was then tested against manually segmented frames using different metrics, intersection over union (IoU), and Boundary F1 (BF) score, and its performance was assessed on various phonatory events on the HSV sequence. The designed network was successfully trained using the hybrid approach, without the need for manual labeling, and tested on the manually labeled data. The performance metrics showed a mean IoU of 0.82 and a mean BF score of 0.96. In addition, the evaluation assessment of the network's performance demonstrated an accurate segmentation of the glottal edges/area even during complex nonstationary phonatory events and when vocal folds were not vibrating, thus overcoming the limitations of the previous hybrid approach that could only be applied to the vibrating vocal folds. The introduced automated scheme guarantees accurate glottis representation in challenging color HSV data with lower image quality and excessive laryngeal maneuvers during all instances of connected speech. This facilitates the future development of HSV-based measures to assess the running vibratory characteristics of the vocal folds in speakers with and without voice disorder. https://doi.org/10.23641/asha.19798864.

Prediction of glioma-subtypes: comparison of performance on a DL classifier using bounding box areas versus annotated tumors

BackgroundFor brain tumors, identifying the molecular subtypes from magnetic resonance imaging (MRI) is desirable, but remains a challenging task. Recent machine learning and deep learning (DL) approaches may help the classification/prediction of tumor subtypes through MRIs. However, most of these methods require annotated data with ground truth (GT) tumor areas manually drawn by medical experts. The manual annotation is a time consuming process with high demand on medical personnel. As an alternative automatic segmentation is often used. However, it does not guarantee the quality and could lead to improper or failed segmented boundaries due to differences in MRI acquisition parameters across imaging centers, as segmentation is an ill-defined problem. Analogous to visual object tracking and classification, this paper shifts the paradigm by training a classifier using tumor bounding box areas in MR images. The aim of our study is to see whether it is possible to replace GT tumor areas by tumor bounding box areas (e.g. ellipse shaped boxes) for classification without a significant drop in performance.MethodIn patients with diffuse gliomas, training a deep learning classifier for subtype prediction by employing tumor regions of interest (ROIs) using ellipse bounding box versus manual annotated data. Experiments were conducted on two datasets (US and TCGA) consisting of multi-modality MRI scans where the US dataset contained patients with diffuse low-grade gliomas (dLGG) exclusively.ResultsPrediction rates were obtained on 2 test datasets: 69.86% for 1p/19q codeletion status on US dataset and 79.50% for IDH mutation/wild-type on TCGA dataset. Comparisons with that of using annotated GT tumor data for training showed an average of 3.0% degradation (2.92% for 1p/19q codeletion status and 3.23% for IDH genotype).ConclusionUsing tumor ROIs, i.e., ellipse bounding box tumor areas to replace annotated GT tumor areas for training a deep learning scheme, cause only a modest decline in performance in terms of subtype prediction. With more data that can be made available, this may be a reasonable trade-off where decline in performance may be counteracted with more data.

Direct experimental evidence for free-space fractional optical vortex transmutation

The emergence of vortex transmutation has opened new ways for vorticity modulation of optical vortices. Although several approaches have been proposed to realize vortex transmutation, fractional optical vortex (FOV) transmutation remains elusive owing to a lack of effective generation and detection methods. Here we report quantitative experimental evidence for a free-space FOV transmutation rule. The key idea is to combine the advantages of a single optical element, termed as fractional spiral polygonal lenses (FSPLs), with a deep learning approach. The desired wavefront is simultaneously generated and manipulated at the focal plane of the FSPL, and the fractional output vorticity is measured by analyzing a single far-field diffraction pattern. Especially, a deep learning scheme using a Bayesian optimization method is developed for output vorticity prediction with a data recovery rate up to 98.2%. The average error of recognized fractional orbital angular momentum modes is as small as 0.02. We clearly observe the intriguing phenomenon that the central vorticity of FOV is changed following a modulo-n transmutation rule in free space. Our results have important implications for fundamental understanding of FOV systems in free space, and offer a technological foundation for potential applications such as quantum information processing and particle manipulation and transportation.

Investigating the Combination of Deep Learning for Channel Estimation and Power Optimization in a Non-Orthogonal Multiple Access System.

In a non-orthogonal multiple access (NOMA) system, the successive interference cancellation (SIC) procedure is typically employed at the receiver side, where several user’s signals are decoded in a subsequent manner. Fading channels may disperse the transmitted signal and originate dependencies among its samples, which may affect the channel estimation procedure and consequently affect the SIC process and signal detection accuracy. In this work, the impact of Deep Neural Network (DNN) in explicitly estimating the channel coefficients for each user in NOMA cell is investigated in both Rayleigh and Rician fading channels. The proposed approach integrates the Long Short-Term Memory (LSTM) network into the NOMA system where this LSTM network is utilized to predict the channel coefficients. DNN is trained using different channel statistics and then utilized to predict the desired channel parameters that will be exploited by the receiver to retrieve the original data. Furthermore, this work examines how the channel estimation based on Deep Learning (DL) and power optimization scheme are jointly utilized for multiuser (MU) recognition in downlink Power Domain Non-Orthogonal Multiple Access (PD-NOMA) system. Power factors are optimized with a view to maximize the sum rate of the users on the basis of entire power transmitted and Quality of service (QoS) constraints. An investigation for the optimization problem is given where Lagrange function and Karush–Kuhn–Tucker (KKT) optimality conditions are applied to deduce the optimum power coefficients. Simulation results for different metrics, such as bit error rate (BER), sum rate, outage probability and individual user capacity, have proved the superiority of the proposed DL-based channel estimation over conventional NOMA approach. Additionally, the performance of optimized power scheme and fixed power scheme are evaluated when DL-based channel estimation is implemented.

Average AoI Minimization in UAV-Assisted Data Collection With RF Wireless Power Transfer: A Deep Reinforcement Learning Scheme

This article studies the unmanned aerial vehicle (UAV)-assisted wireless powered network, where a UAV is dispatched to wirelessly charge multiple ground nodes (GNs) by using radio frequency (RF) energy transfer and then the GNs use their harvested energy to upload the sensed information to the UAV. At each moment, the UAV is scheduled to charge the GNs or only one GN is scheduled to upload its data. An optimization problem is formulated to minimize the average Age of Information (AoI) of the GNs by jointly optimizing the trajectory of the UAV and the scheduling of information transmission and energy harvesting of GNs. As the problem is a combinational optimization problem with a set of binary variables, it is difficult to be solved. Thus, it is modeled as a Markov problem with large state spaces and a deep <inline-formula> <tex-math notation="LaTeX">${Q}$ </tex-math></inline-formula> network (DQN)-based scheme is proposed to find its near-optimal solution on the basis of the deep reinforcement learning (DRL) framework. Two nets are structured with artificial neural network (ANN), where one is for evaluating the reward of the action performed in current state, and the other is for predicting realistic action. The corresponding state spaces, the efficient action spaces, and reward function are designed. Simulation results demonstrate the convergence of the proposed DQN scheme, which also show that the proposed DQN scheme gets much smaller average AoI than the three other known schemes. Moreover, by involving the energy punishment in the reward, the UAV may save its energy but yield higher AoI. Additionally, the effects of the packet size, the transmit power, and the distribution area of GNs on the GNs&#x2019; average AoI are also discussed, which are expected to provide some useful insights.

A Large-Area Flexible Tactile Sensor for Multi-Touch and Force Detection Using Electrical Impedance Tomography

For large area robot skin design, the distribution of rigid components and wires in traditional array sensors leads to the decrease of the flexibility and extensibility of sensors. The flexible sensors based on non-invasive electrical impedance tomography (EIT) can avoid these shortcomings. However, design of a large-area flexible tactile sensors using EIT technique is still challenging due to the tradeoff between manufacturability and spatial resolution. In this study, we proposed a novel non-array EIT-based tactile sensor which is made of a porous elastic polymer and ionic liquid. The sensor free from internal array electrodes is straightforward to manufacture and can cover a large area with low cost. To improve the spatial resolution and touch sensitivity, we adopted a deep learning scheme Pyramid Scene Parsing Network (PSPNet) to postprocess the originally reconstructed images to enhance the sensor tactile perception. With this data-driven method, we achieved a single-point position detection error of 7.5± 4.5 mm without using internal electrodes. To overcome the location dependency of EIT sensing problems, we proposed a method of sub-regional fitting to calibrate the distributed forces for the large-area flexible tactile sensor and obtained a quantitative relationship between the touch force and EIT measurement for the entire sensing area for continuous sensing. The real-time performance of the proposed prototype sensor system demonstrated that we can achieve more accuracy of multi-touch and distributed force detection on the non-array sensors with a temporal resolution of about 1.3 frames/sec for potential applications in human-robot interfaces.