Thesis Comparación de algoritmos basados en deep-learning para clasificación de potenciales evocados P300 en EEG
No Thumbnail Available
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Program
Ingeniería Civil Informática
Departament
Campus
Campus Santiago San Joaquín
Abstract
Las interfaces cerebro-computador (BCI, en inglés) son sistemas de control y comunicación que buscan traducir actividad cerebral en comandos para controlar diversos dispositivos. La técnica más popular para registrar señales cerebrales es la electroencefalografía (EEG), que permite la detección de potenciales evocados. En los últimos años, se han comenzado a implementar modelos de Deep Learning para la detección de potenciales evocados. Sin embargo, debido a la novedad de estas soluciones, aún no se ha realizado una comparación entre ellas. Por lo tanto, se propone comparar con una metodología estándar, diferentes modelos de Deep Learning propuestos en los últimos años.
Brain-computer interfaces (BCI) are control and communication systems that seek to translate brain activity into commands to control various devices. The most popular technique for recording brain signals is electroencephalography (EEG), which allows the detection of evoked potentials. In recent years, Deep Learning models have begun to be implemented for the detection of evoked potentials. However, due to the novelty of these solutions, a comparison between them has not yet been made. Therefore, it is proposed to compare with a standard methodology, different Deep Learning models proposed in the last years.
Brain-computer interfaces (BCI) are control and communication systems that seek to translate brain activity into commands to control various devices. The most popular technique for recording brain signals is electroencephalography (EEG), which allows the detection of evoked potentials. In recent years, Deep Learning models have begun to be implemented for the detection of evoked potentials. However, due to the novelty of these solutions, a comparison between them has not yet been made. Therefore, it is proposed to compare with a standard methodology, different Deep Learning models proposed in the last years.
Description
Keywords
Interfaces cerebro-computador, P300 Speller, Aprendizaje profundo
