SEMG-Based Hand Posture Recognition Considering Electrode Shift, Feature Vectors, and Posture Groups.
Ontology highlight
ABSTRACT:
SUBMITTER: Kim J
PROVIDER: S-EPMC8624257 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Hand gesture recognition tasks based on surface electromyography (sEMG) are vital in human-computer interaction, speech detection, robot control, and rehabilitation applications. However, existing models, whether traditional machine learnings (ML) or other state-of-the-arts, are limited in the number of movements. Targeting a large number of gesture classes, more data features such as temporal information should be persisted as much as possible. In the field of sEMG-based recognitions, the recurrent convolutional neural network (RCNN) is an advanced method due to the sequential characteristic of sEMG signals. However, the invariance of the pooling layer damages important temporal information. In the all convolutional neural network (ACNN), because of the feature-mixing convolution operation, a same output can be received from completely different inputs. This paper proposes a concatenate feature fusion (CFF) strategy and a novel concatenate feature fusion recurrent convolutional neural network (CFF-RCNN). In CFF-RCNN, a max-pooling layer and a 2-stride convolutional layer are concatenated together to replace the conventional simple dimensionality reduction layer. The featurewise pooling operation serves as a signal amplitude detector without using any parameter. The feature-mixing convolution operation calculates the contextual information. Complete evaluations are made on both the accuracy and convergence speed of the CFF-RCNN. Experiments are conducted using three sEMG benchmark databases named DB1, DB2 and DB4 from the NinaPro database. With more than 50 gestures, the classification accuracies of the CFF-RCNN are 88.87% on DB1, 99.51% on DB2, and 99.29% on DB4. These accuracies are the highest compared with reported accuracies of machine learnings and other state-of-the-art methods. To achieve accuracies of 86%, 99% and 98% for the RCNN, the training time are 2353.686 s, 816.173 s and 731.771 s, respectively. However, for the CFF-RCNN to reach the same accuracies, it needs only 1727.415 s, 542.245 s and 576.734 s, corresponding to a reduction of 26.61%, 33.56% and 21.19% in training time. We concluded that the CFF-RCNN is an improved method when classifying a large number of hand gestures. The CFF strategy significantly improved model performance with higher accuracy and faster convergence as compared to traditional RCNN.
| S-EPMC8775254 | biostudies-literature
Project description:Surface electromyography (sEMG) sensors are widely used in the fields of ergonomics, sports science, and medical research. However, current sEMG sensors cannot recognize the various exercise intensities efficiently because of the strain interference, low conductivity, and poor skin-conformability of their electrodes. Here, we present a highly conductive, strain-insensitive, and low electrode-skin impedance elastic sEMG electrode, which consists of a three-layered structure (polydimethylsiloxane/galinstan + polydimethylsiloxane/silver-coated nickel + polydimethylsiloxane). The bottom layer of the electrode consists of vertically conductive magnetic particle paths, which are insensitive to stretching strain, collect sEMG charge from human skin, and finally transfer it to processing circuits via an intermediate layer. Our skin-friendly electrode exhibits high conductivity (0.237 and 1.635 mΩ.cm resistivities in transverse and longitudinal directions, respectively), low electrode-skin impedance (47.23 kΩ at 150 Hz), excellent strain-insensitivity (10% change of electrode-skin impedance within the 0%-25% strain range), high fatigue resistance (>1500 cycles), and good conformability with skin. During various exercise intensities, the signal-to-noise ratio (SNR) of our electrode increased by 22.53 dB, which is 206% and 330% more than that of traditional Ag/AgCl and copper electrode, respectively. The ability of our electrode to efficiently recognize various exercise intensities confirms its great application potential for the field of sports health.
| S-EPMC7143104 | biostudies-literature
Project description:We have developed a novel algorithm for sEMG feature extraction and classification. It is based on a hybrid network composed of spiking and artificial neurons. The spiking neuron layer with mutual inhibition was assigned as feature extractor. We demonstrate that the classification accuracy of the proposed model could reach high values comparable with existing sEMG interface systems. Moreover, the algorithm sensibility for different sEMG collecting systems characteristics was estimated. Results showed rather equal accuracy, despite a significant sampling rate difference. The proposed algorithm was successfully tested for mobile robot control.
| S-EPMC4701259 | biostudies-literature
Project description:This paper presents an electromyography (EMG) signal dataset for use in human-computer interaction studies. The dataset includes 4-channel surface EMG data from 40 participants with an equal gender distribution. The gestures in the data are rest or neutral state, extension of the wrist, flexion of the wrist, ulnar deviation of the wrist, radial deviation of the wrist, grip, abduction of all fingers, adduction of all fingers, supination, and pronation. Data were collected from 4 forearm muscles when simulating 10 unique hand gestures and recorded with the BIOPAC MP36 device using Ag/AgCl surface bipolar electrodes. Each participant's data contains five repetitive cycles of ten hand gestures. A demographic survey was applied to the participants before the signal recording process. This data can be utilized for recognition, classification, and prediction studies in order to develop EMG-based hand movement controller systems. The dataset can also be useful as a reference to create an artificial intelligence model (especially a deep learning model) to detect gesture-related EMG signals. Additionally, it is encouraged to use the proposed dataset for benchmarking current datasets in the literature or for validation of machine learning and deep learning models created with different datasets in accordance with the participant-independent validation strategy.
| S-EPMC8844426 | biostudies-literature
Project description:Grasp force estimation based on surface electromyography (sEMG) is essential for the dexterous control of a prosthetic hand. Nowadays, although increasing the number of sEMG measurement positions and extracting more features are common methods to increase the accuracy of grasp force estimation, it will increase the computational burden. In this paper, an approach based on analysis of variance (ANOVA) and generalized regression neural network (GRNN) for optimal measurement positions and features is proposed, with the purpose of using fewer measurement positions or features to achieve higher estimation accuracy. Firstly, we captured six channels of sEMG from subjects' forearm and grasp force synchronously. Then, four kinds of features in time domain are extracted from each channel of sEMG. By combining different measurement position sets (MPSs) and feature set (FSs), we construct 945 data sets. These data sets are fed to GRNN to realize grasp force estimation. Normalized root mean square error (NRMS), normalized mean of absolute error (NMAE), and correlation coefficient (CC) between estimated grasp force and actual force are introduced to evaluate the performance of grasp force estimation. Finally, ANOVA and Tukey HSD testing are introduced to analyze grasp force estimation results so as to obtain the optimal measurement positions and features. We obtain the optimal MPSs for grasp force estimation when different FSs are employed, and the optimal FSs when different MPSs are utilized.
| S-EPMC8009426 | biostudies-literature
Project description:In lower-limb rehabilitation, human action recognition (HAR) technology can be introduced to analyze the surface electromyography (sEMG) signal generated by movements, which can provide an objective and accurate evaluation of the patient's action. To balance the long cycle required for rehabilitation and the inconvenient factors brought by wearing sEMG devices, a portable sEMG signal acquisition device was developed that can be used under daily scenarios. Additionally, a mobile application was developed to meet the demand for real-time monitoring and analysis of sEMG signals. This application can monitor data in real time and has functions such as plotting, filtering, storage, and action capture and recognition. To build the dataset required for the recognition model, six lower-limb motions were developed for rehabilitation (kick, toe off, heel off, toe off and heel up, step back and kick, and full gait). The sEMG segment and action label were combined for training a convolutional neural network (CNN) to achieve high-precision recognition performance for human lower-limb actions (with a maximum accuracy of 97.96% and recognition accuracy for all actions reaching over 97%). The results show that the smartphone-based sEMG analysis system proposed in this paper can provide reliable information for the clinical evaluation of lower-limb rehabilitation.
| S-EPMC10452551 | biostudies-literature
Project description:Bone is capable of adapting during life in response to stress. Therefore, variation in locomotor and manipulative behaviours across extant hominoids may be reflected in differences in trabecular bone structure. The hand is a promising region for trabecular analysis, as it is the direct contact between the individual and the environment and joint positions at peak loading vary amongst extant hominoids. Building upon traditional volume of interest-based analyses, we apply a whole-epiphysis analytical approach using high-resolution microtomographic scans of the hominoid third metacarpal to investigate whether trabecular structure reflects differences in hand posture and loading in knuckle-walking (Gorilla, Pan), suspensory (Pongo, Hylobates and Symphalangus) and manipulative (Homo) taxa. Additionally, a comparative phylogenetic method was used to analyse rates of evolutionary changes in trabecular parameters. Results demonstrate that trabecular bone volume distribution and regions of greatest stiffness (i.e., Young's modulus) correspond with predicted loading of the hand in each behavioural category. In suspensory and manipulative taxa, regions of high bone volume and greatest stiffness are concentrated on the palmar or distopalmar regions of the metacarpal head, whereas knuckle-walking taxa show greater bone volume and stiffness throughout the head, and particularly in the dorsal region; patterns that correspond with the highest predicted joint reaction forces. Trabecular structure in knuckle-walking taxa is characterised by high bone volume fraction and a high degree of anisotropy in contrast to the suspensory brachiators. Humans, in which the hand is used primarily for manipulation, have a low bone volume fraction and a variable degree of anisotropy. Finally, when trabecular parameters are mapped onto a molecular-based phylogeny, we show that the rates of change in trabecular structure vary across the hominoid clade. Our results support a link between inferred behaviour and trabecular structure in extant hominoids that can be informative for reconstructing behaviour in fossil primates.
| S-EPMC3828321 | biostudies-literature
Project description:BackgroundThe origin of human handedness and its evolution in primates is presently under debate. Current hypotheses suggest that body posture (postural origin hypothesis and bipedalism hypothesis) have an important impact on the evolution of handedness in primates. To gain insight into the origin of manual lateralization in primates, we studied gray mouse lemurs, suggested to represent the most ancestral primate condition. First, we investigated hand preference in a simple food grasping task to explore the importance of hand usage in a natural foraging situation. Second, we explored the influence of body posture by applying a forced food grasping task with varying postural demands (sit, biped, cling, triped).ResultsThe tested mouse lemur population did not prefer to use their hands alone to grasp for food items. Instead, they preferred to pick them up using a mouth-hand combination or the mouth alone. If mouth usage was inhibited, they showed an individual but no population level handedness for all four postural forced food grasping tasks. Additionally, we found no influence of body posture on hand preference in gray mouse lemurs.ConclusionOur results do not support the current theories of primate handedness. Rather, they propose that ecological adaptation indicated by postural habit and body size of a given species has an important impact on hand preference in primates. Our findings suggest that small-bodied, quadrupedal primates, adapted to the fine branch niche of dense forests, prefer mouth retrieval of food and are less manually lateralized than large-bodied species which consume food in a more upright, and less stable body posture.
| S-EPMC3056780 | biostudies-literature
Project description:BackgroundThe human hand can perform a range of manipulation tasks, from holding a pen to holding a hammer. The central nervous system (CNS) uses different strategies in different manipulation tasks based on task requirements. Attempts to compare postures of the hand have been made for use in robotics and animation industries. In this study, we developed an index called the posture similarity index to quantify the similarity between two human hand postures.MethodsTwelve right-handed volunteers performed 70 postures, and lifted and held 30 objects (total of 100 different postures, each performed five times). A 16-sensor electromagnetic tracking system captured the kinematics of individual finger phalanges (segments). We modeled the hand as a 21-DoF system and computed the corresponding joint angles. We used principal component analysis to extract kinematic synergies from this 21-DoF data. We developed a posture similarity index (PSI), that represents the similarity between posture in the synergy (Principal component) space. First, we tested the performance of this index using a synthetic dataset. After confirming that it performs well with the synthetic dataset, we used it to analyze the experimental data. Further, we used PSI to identify postures that are "representative" in the sense that they have a greater overlap (in synergy space) with a large number of postures.ResultsOur results confirmed that PSI is a relatively accurate index of similarity in synergy space both with synthetic data and real experimental data. Also, more special postures than common postures were found among "representative" postures.ConclusionWe developed an index for comparing posture similarity in synergy space and demonstrated its utility by using synthetic dataset and experimental dataset. Besides, we found that "special" postures are actually "special" in the sense that there are more of them in the "representative" postures as identified by our posture similarity index.
| S-EPMC6292379 | biostudies-literature
Project description:The filtering feature-selection algorithm is a kind of important approach to dimensionality reduction in the field of the text categorization. Most of filtering feature-selection algorithms evaluate the significance of a feature for category based on balanced dataset and do not consider the imbalance factor of dataset. In this paper, a new scheme was proposed, which can weaken the adverse effect caused by the imbalance factor in the corpus. We evaluated the improved versions of nine well-known feature-selection methods (Information Gain, Chi statistic, Document Frequency, Orthogonal Centroid Feature Selection, DIA association factor, Comprehensive Measurement Feature Selection, Deviation from Poisson Feature Selection, improved Gini index, and Mutual Information) using naïve Bayes and support vector machines on three benchmark document collections (20-Newsgroups, Reuters-21578, and WebKB). The experimental results show that the improved scheme can significantly enhance the performance of the feature-selection methods.
| S-EPMC4058251 | biostudies-other