Descripción
|
|
---|---|
Modern actuated prostheses for upper-limb loss patients provide many degrees of freedom (DOF) and mimic natural limbs well, but robust, multi-DOF control of such devices has not yet been achieved. Electromyography (EMG) signals based on patients? intact muscle activity may be used as control signals, but most available algorithms employ very simple encoding and mapping of EMG features to actuation of few DOFs. Modern machine learning methods can be used to classify gestures and movements from EMG features, but their accuracy is degraded by variance in signal properties due to changing electrode placement, arm position, and other contextual variations. | |
Internacional
|
Si |
Nombre congreso
|
Neuroscience 2018 |
Tipo de participación
|
970 |
Lugar del congreso
|
San Diego |
Revisores
|
Si |
ISBN o ISSN
|
978-0-19-530548-7 |
DOI
|
|
Fecha inicio congreso
|
03/11/2018 |
Fecha fin congreso
|
07/11/2018 |
Desde la página
|
1 |
Hasta la página
|
1 |
Título de las actas
|
Neuroscience |