Human-in-the-loop optimization (HILO) has been used to identify subject-specific assistive strategies and improve the performance of wearable exoskeletons. However, there is still a gap in research on HILO regarding knee exoskeleton flexion assistance. We present methodology that optimizes torque delivered by exoskeleton. The cooperation mechanism flexor antagonist muscles assisted first analyzed through dynamic model. Furthermore, online rapid metabolic evaluation function, including...

The detection of moving objects in videos is very important many video processing applications, and background modeling often an indispensable process to achieve this goal. Most the traditional methods utilize color or texture information. However, information sensitive illumination variations cannot be utilized separate smooth foreground from most cases. Achieving good performance terms high accuracy low computational cost also challenging. In paper, we propose a new integration framework...

Both the discrete motion states and continuous joint kinematics are essential for controlling assistive robots under changeable environmental conditions. However, few studies investigate both intents. This article is first work to propose an end-to-end intent decoding method that integrates recognition of locomotion modes prediction gait events. First, we a data-driven approach segment transitional periods adjacent determine their boundaries. Second, build convolutional neural network...

The interaction between human and exoskeletons increasingly relies on the precise decoding of motion. One main issue current motion algorithms is that seldom provide both discrete patterns (e.g., gait phases) continuous parameters kinematics). In this paper, we propose a novel algorithm uses surface electromyography (sEMG) signals are generated prior to their corresponding motions perform phase recognition lower-limb kinematics prediction. Particularly, first an end-to-end architecture EMG...

The cybertwin-driven 6G that can obtain static and dynamic data stream of users provide an exciting potential for a novel muscular human cybertwin beyond traditonally used artificial neural networks (ANNs) musculoskeletal models (MSMs). In this article, we propose the conceptual design construct baseline model with improved generalization ability over ANN easier adaptation to new distributions MSMs. particular, first time combine MSM, which benefits from combination learning-based approaches...

Deep learning (DL)-based human activity recognition (HAR) methods have shown promise in the applications of health Internet Things (IoT) and wireless body sensor networks (BSNs). However, adapting these to new users real-world scenarios is challenging due cross-subject issue. To solve this issue, we propose ActiveSelfHAR, a framework that combines active learning's benefit sparsely acquiring informative samples with actual labels self-training's effectively utilizing unlabeled data adapt HAR...

This paper presents an occlusion robust image representation method and apply it to face recognition. Inspired from the recent work [15], we propose a Gabor phase difference for Based on good ability of filters capture structure robustness shown in this paper, features are expected be discriminative case. Besides, adopt spectral regression based discriminant analysis with extracted find most subspace classify different faces. In way, is derived. Extensive experiments various cases show...

Surface Electromyography (sEMG) enables an intuitive control of wearable robots. The muscle fatigue-induced changes sEMG signals might limit the long-term usage sEMG-based algorithms. This paper presents performance deterioration gait phase classifiers, explains by analyzing time-varying extracted features, and proposes a training strategy that can improve classifiers’ robustness against fatigue. In particular, we first select some features are commonly used in fatigue-related studies use...

The prevalence of smart devices encourages increasing requirements wearable human–computer interactions. To improve user acceptance, such interactions require easy-to-manipulate and unobtrusive characteristics. In this article, we, for the first time, propose to recognize silent commands through a lightweight around-ear biosensing system Mordo that can be easily integrated with earphones, manipulate devices, minimize social awkwardness. particular, we determine empirical principles...

An occlusion robust image representation method is presented and applied to face recognition. In our method, Gabor phase difference used mainly resist occlusion. Based on the good ability of filters capture structure robustness shown here, features are expected be discriminative for in case. Furthermore, we find that different scales orientations lead quite varied performance then analyze it carefully effective (EGP) features. Moreover, adopt spectral regression–based discriminant analysis,...

Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI

Traditional load prediction methods are unable to effectively predict the loads according spatial topology of each electricity consumer in neighboring areas and dependency correlations. In order further improve accuracy region, this paper proposes a short-term method electric based on multi-graph convolutional network. First, input data selected with maximum information coefficient by integrating multi-dimensional such as load, weather, price date areas. Then, gated network is used temporal...

Brain decoding that classifies cognitive states using the functional fluctuations of brain can provide insightful information for understanding mechanisms functions. Among common procedures with magnetic resonance imaging (fMRI), extracting time series each region after parcellation traditionally averages across voxels within a region. This neglects spatial among and requirement downstream tasks. In this study, we propose to use fully connected neural network is jointly trained decoder...

Brain decoding that classifies cognitive states using the functional fluctuations of brain can provide insightful information for understanding mechanisms functions. Among common procedures with magnetic resonance imaging (fMRI), extracting time series each region after parcellation traditionally averages across voxels within a region. This neglects spatial among and requirement downstream tasks. In this study, we propose to use fully connected neural network is jointly trained decoder...

This article is based on deep learning algorithms and uses MRI to study the development of congenital heart septal defects in neonatal brain tissue.From January 2018 December 2019, 150 cases cardiac paper defect were retrospectively analyzed 50 normal newborns neonates. The four index parametersbrain MR imaging, lateral ventricle pre-angle measurement (F/F'), body (D/ D'), caudal nucleus (C/C') analyzed. independent sample t test performed compare difference parameters between groups.F...

The surface electromyography (sEMG) signal-based human-machine interface (HMI) has been widely used for various scenarios of physical human-robot interaction. However, current HMIs based on bipolar myoelectric sensors are hindered by the limitations global sEMG features, which prone to variability and delay. In this letter, we define a HMI that takes advantage underlying neural information spinal module activations from signals, inspired recent findings codes. Firstly, identified spiking...

Wearable human-computer interactions in daily life are increasingly encouraged by the prevalence of intelligent wearables. It poses a demanding requirement micro-interaction and minimizing social awkwardness. Our previous work demonstrated feasibility recognizing silent commands through around-ear biosensors with limitation user adaptation. In this work, we ease personalization framework that integrates spectral factorization signals, temporal confidence rejection commonly used transfer...