Integrated Word2Vec-based Speech Annotations for Enhanced EEG Decoding of Speech Intentions

Se Na Jang; Hyeong Yeong Park; Seong Wook Kim; Seong Hyun Yu; Minji Lee; Ji Hoon Jeong

doi:10.1109/BCI65088.2025.10931508

Integrated Word2Vec-based Speech Annotations for Enhanced EEG Decoding of Speech Intentions

Se Na Jang
, Hyeong Yeong Park
, Seong Wook Kim
, Seong Hyun Yu
, Minji Lee
, Ji Hoon Jeong

Chungbuk National University

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

2 Scopus citations

Abstract

Brain-computer interface (BCI) technology has promising applications as an intuitive communication tool and in fields such as language rehabilitation. This study aims to decode human speech intentions by analyzing EEG signals recorded during actual and imagined speech. EEG data were collected using a 64 channels system, with preprocessing to remove artifacts. Automatic speech recognition (ASR) was used to extract precise speech onset times, generating time-specific speech annotations corresponding with EEG data. Pretrained Word2Vec embeddings were integrated to provide semantic context, combining neural signals with high-level linguistic features. Support vector machine (SVM), linear discriminant analysis (LDA) were employed for decoding. The results demonstrate that integrating speech annotations improves decoding accuracy, even for imagined speech, highlighting the potential of BCI technology for advanced applications in communication and rehabilitation.

Original language	English
Title of host publication	13th International Winter Conference on Brain-Computer Interface, BCI 2025
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9798331521929
DOIs	https://doi.org/10.1109/BCI65088.2025.10931508
State	Published - 2025
Event	13th International Winter Conference on Brain-Computer Interface, BCI 2025 - Hybrid, Gangwon, Korea, Republic of Duration: 24 Feb 2025 → 26 Feb 2025

Publication series

Name	International Winter Conference on Brain-Computer Interface, BCI
ISSN (Print)	2572-7672

Conference

Conference	13th International Winter Conference on Brain-Computer Interface, BCI 2025
Country/Territory	Korea, Republic of
City	Hybrid, Gangwon
Period	24/02/25 → 26/02/25

Bibliographical note

Publisher Copyright:
© 2025 IEEE.

Keywords

Automatic speech recognition
EEG
Sentence decoding
Word2Vec

Access to Document

10.1109/BCI65088.2025.10931508

Cite this

Jang, S. N., Park, H. Y., Kim, S. W., Yu, S. H., Lee, M., & Jeong, J. H. (2025). Integrated Word2Vec-based Speech Annotations for Enhanced EEG Decoding of Speech Intentions. In 13th International Winter Conference on Brain-Computer Interface, BCI 2025 (International Winter Conference on Brain-Computer Interface, BCI). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/BCI65088.2025.10931508

@inproceedings{14b0f579b45e4558b49af07173e57b70,

title = "Integrated Word2Vec-based Speech Annotations for Enhanced EEG Decoding of Speech Intentions",

abstract = "Brain-computer interface (BCI) technology has promising applications as an intuitive communication tool and in fields such as language rehabilitation. This study aims to decode human speech intentions by analyzing EEG signals recorded during actual and imagined speech. EEG data were collected using a 64 channels system, with preprocessing to remove artifacts. Automatic speech recognition (ASR) was used to extract precise speech onset times, generating time-specific speech annotations corresponding with EEG data. Pretrained Word2Vec embeddings were integrated to provide semantic context, combining neural signals with high-level linguistic features. Support vector machine (SVM), linear discriminant analysis (LDA) were employed for decoding. The results demonstrate that integrating speech annotations improves decoding accuracy, even for imagined speech, highlighting the potential of BCI technology for advanced applications in communication and rehabilitation.",

keywords = "Automatic speech recognition, EEG, Sentence decoding, Word2Vec",

author = "Jang, \{Se Na\} and Park, \{Hyeong Yeong\} and Kim, \{Seong Wook\} and Yu, \{Seong Hyun\} and Minji Lee and Jeong, \{Ji Hoon\}",

note = "Publisher Copyright: {\textcopyright} 2025 IEEE.; 13th International Winter Conference on Brain-Computer Interface, BCI 2025 ; Conference date: 24-02-2025 Through 26-02-2025",

year = "2025",

doi = "10.1109/BCI65088.2025.10931508",

language = "English",

series = "International Winter Conference on Brain-Computer Interface, BCI",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "13th International Winter Conference on Brain-Computer Interface, BCI 2025",

address = "United States",

}

Jang, SN, Park, HY, Kim, SW, Yu, SH, Lee, M & Jeong, JH 2025, Integrated Word2Vec-based Speech Annotations for Enhanced EEG Decoding of Speech Intentions. in 13th International Winter Conference on Brain-Computer Interface, BCI 2025. International Winter Conference on Brain-Computer Interface, BCI, Institute of Electrical and Electronics Engineers Inc., 13th International Winter Conference on Brain-Computer Interface, BCI 2025, Hybrid, Gangwon, Korea, Republic of, 24/02/25. https://doi.org/10.1109/BCI65088.2025.10931508

Integrated Word2Vec-based Speech Annotations for Enhanced EEG Decoding of Speech Intentions. / Jang, Se Na; Park, Hyeong Yeong; Kim, Seong Wook et al.
13th International Winter Conference on Brain-Computer Interface, BCI 2025. Institute of Electrical and Electronics Engineers Inc., 2025. (International Winter Conference on Brain-Computer Interface, BCI).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Integrated Word2Vec-based Speech Annotations for Enhanced EEG Decoding of Speech Intentions

AU - Jang, Se Na

AU - Park, Hyeong Yeong

AU - Kim, Seong Wook

AU - Yu, Seong Hyun

AU - Lee, Minji

AU - Jeong, Ji Hoon

PY - 2025

Y1 - 2025

N2 - Brain-computer interface (BCI) technology has promising applications as an intuitive communication tool and in fields such as language rehabilitation. This study aims to decode human speech intentions by analyzing EEG signals recorded during actual and imagined speech. EEG data were collected using a 64 channels system, with preprocessing to remove artifacts. Automatic speech recognition (ASR) was used to extract precise speech onset times, generating time-specific speech annotations corresponding with EEG data. Pretrained Word2Vec embeddings were integrated to provide semantic context, combining neural signals with high-level linguistic features. Support vector machine (SVM), linear discriminant analysis (LDA) were employed for decoding. The results demonstrate that integrating speech annotations improves decoding accuracy, even for imagined speech, highlighting the potential of BCI technology for advanced applications in communication and rehabilitation.

AB - Brain-computer interface (BCI) technology has promising applications as an intuitive communication tool and in fields such as language rehabilitation. This study aims to decode human speech intentions by analyzing EEG signals recorded during actual and imagined speech. EEG data were collected using a 64 channels system, with preprocessing to remove artifacts. Automatic speech recognition (ASR) was used to extract precise speech onset times, generating time-specific speech annotations corresponding with EEG data. Pretrained Word2Vec embeddings were integrated to provide semantic context, combining neural signals with high-level linguistic features. Support vector machine (SVM), linear discriminant analysis (LDA) were employed for decoding. The results demonstrate that integrating speech annotations improves decoding accuracy, even for imagined speech, highlighting the potential of BCI technology for advanced applications in communication and rehabilitation.

KW - Automatic speech recognition

KW - EEG

KW - Sentence decoding

KW - Word2Vec

UR - https://www.scopus.com/pages/publications/105002342624

U2 - 10.1109/BCI65088.2025.10931508

DO - 10.1109/BCI65088.2025.10931508

M3 - Conference contribution

AN - SCOPUS:105002342624

T3 - International Winter Conference on Brain-Computer Interface, BCI

BT - 13th International Winter Conference on Brain-Computer Interface, BCI 2025

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 13th International Winter Conference on Brain-Computer Interface, BCI 2025

Y2 - 24 February 2025 through 26 February 2025

ER -

Jang SN, Park HY, Kim SW, Yu SH, Lee M, Jeong JH. Integrated Word2Vec-based Speech Annotations for Enhanced EEG Decoding of Speech Intentions. In 13th International Winter Conference on Brain-Computer Interface, BCI 2025. Institute of Electrical and Electronics Engineers Inc. 2025. (International Winter Conference on Brain-Computer Interface, BCI). doi: 10.1109/BCI65088.2025.10931508

Integrated Word2Vec-based Speech Annotations for Enhanced EEG Decoding of Speech Intentions

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this