témavezető: Al-Radhi Mohammed Salah
helyszín (magyar oldal): Department of Telecommunications And Media Informatics helyszín rövidítés: TMIT
A kutatási téma leírása:
This thesis proposal aims to advance the field of Brain-Computer Interfaces (BCIs) by using deep learning techniques for decoding speech envelope from brain signals, including EEG and ECoG. The study will focus on exploring the ability of deep learning models to enhance the decoding of speech features from diverse brain signals, with a particular emphasis on their application in BCI technologies. Also, the research will investigate the generalization capabilities of deep learning models across different types of brain signals and their implications for real-world BCI applications (such as speech synthesis).
Objectives:
1. Investigate the application of deep learning models in decoding speech envelopes from EEG and ECoG signals for improved accuracy and efficiency in BCI systems.
2. Explore the unique characteristics of EEG and ECoG signals and their impact on the performance of deep learning architectures for speech decoding.
3. Assess the generalization capabilities of deep learning models for decoding speech features from various brain signals, including EEG and ECoG.
Methodology:
1. Implement deep learning architectures shaped for decoding speech envelope from EEG and ECoG signals using appropriate datasets.
2. Compare the performance of deep learning models with traditional linear models in decoding speech features from different types of brain signals.
3. Conduct experiments to evaluate the generalization capabilities of deep learning models across diverse brain signal modalities.
4. Discuss the results in the context of existing literature on deep learning applications in neuroscience and the development of advanced BCI systems.
5. Disseminate research findings through academic publications, presentations at relevant conferences.
Potential Impact:
The findings may pave the way for more efficient and accurate BCI systems that can benefit individuals with communication impairments and motor disabilities.
References:
Accou, B., Vanthornhout, J., Van hamme, H., & Francart, T. (2023). Decoding of the speech envelope from EEG using deep neural networks. Nature Scientific Reports, 13:812. https://doi.org/10.1038/s41598-022-27332-2
előírt nyelvtudás: English felvehető hallgatók száma: 1
Jelentkezési határidő: 2024-06-19
2024. IV. 17. ODT ülés Az ODT következő ülésére 2024. június 14-én, pénteken 10.00 órakor kerül sor a Semmelweis Egyetem Szenátusi termében (Bp. Üllői út 26. I. emelet).
Bejelentkezés
