http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Analysis of biosignal using artificial neural networks for human-machine interface
Woochul Nam(남우철) 대한기계학회 2021 대한기계학회 춘추학술대회 Vol.2021 No.4
Robust human-machine interface (HMI) can be realized with accurate estimations on user intention (or motion) using his/her biosignal. A model-based approach is inefficient to study biosignal because the data is affected by numerous factors. Thus, artificial neural networks can be an effective computational tool for HMI. Although neural network-based approach is able to estimate states of HMI users with their biosignals, classical neural networks could show limited performances. If stressful experiments are required to acquire biosignals, it is difficult to obtain sufficient datasets for deep neural networks. Moreover, most biosignals show large variations between HMI users. To resolve these issues, data synthesis approaches can be used because they are able to generate artificial biosignal, which increases the size of the datasets. Furthermore, the performance degradation owing to cross-subject variability can be addressed by using domain adaptation neural networks. This work introduces various methods for data synthesis and domain adaptation for HMI.