IEICE Technical Committee Submission System
Conference Paper's Information
Online Proceedings
[Sign in]
Tech. Rep. Archives
 Go Top Page Go Previous   [Japanese] / [English] 

Paper Abstract and Keywords
Presentation 2021-03-04 16:10
Estimation of imagined speech from electrocorticogram with an encoder-decoder model
Kotaro Hayashi, Shuji Komeiji (TUAT), Takumi Mitsuhashi, Yasushi Iimura, Hiroharu Suzuki, Hidenori Sugano (Juntendo Univ.), Koichi Shinoda (TokyoTech), Toshihisa Tanaka (TUAT) EA2020-87 SIP2020-118 SP2020-52
Abstract (in Japanese) (See Japanese page) 
(in English) Recent advances in signal processing and machine learning technologies have made it possible to estimate and reconstruct speech or text during speaking and listening from invasive electrocorticogram (ECoG). Meanwhile, the estimation of imagined speech has not been successful due to the difficulty in synchronizing the ECoG with the target label. In this paper, we hypothesize that if imagined speech and ECoG are adequately synchronized, speech can be decoded from ECoG as in the case of ECoG during speaking and listening. We designed an experiment in which participants were presented with a screen on which short sentences were projected, and by highlighting the colors of the letters one by one, the timing and speed of imagination could be controlled. The ECoG was recorded during the three tasks of imagining speech, listening to speech, and speaking. Moreover, we recorded the speech presented in the listening task and the participants' speech in the speaking task. From the measured ECoG, we built an encoder-decoder model using the Mel cepstrum coefficients of the speech during the speaking and listening tasks to infer the imagined speech. The best character error rate of about 17% was achieved for sentences decoded from the imagined ECoG.
Keyword (in Japanese) (See Japanese page) 
(in English) Electrocorticogram (ECoG) / Imaginary speech / Brain-computer interface (BCI) / Encoder-decoder model / / / /  
Reference Info. IEICE Tech. Rep., vol. 120, no. 398, SIP2020-118, pp. 164-169, March 2021.
Paper # SIP2020-118 
Date of Issue 2021-02-24 (EA, SIP, SP) 
ISSN Print edition: ISSN 0913-5685  Online edition: ISSN 2432-6380
Copyright
and
reproduction
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034)
Download PDF EA2020-87 SIP2020-118 SP2020-52

Conference Information
Committee EA US SP SIP IPSJ-SLP  
Conference Date 2021-03-03 - 2021-03-04 
Place (in Japanese) (See Japanese page) 
Place (in English) Online 
Topics (in Japanese) (See Japanese page) 
Topics (in English) Speech, Engineering/Electro Acoustics, Signal Processing, Ultrasonics, and Related Topics 
Paper Information
Registration To SIP 
Conference Code 2021-03-EA-US-SP-SIP-SLP 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Estimation of imagined speech from electrocorticogram with an encoder-decoder model 
Sub Title (in English)  
Keyword(1) Electrocorticogram (ECoG)  
Keyword(2) Imaginary speech  
Keyword(3) Brain-computer interface (BCI)  
Keyword(4) Encoder-decoder model  
Keyword(5)  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Kotaro Hayashi  
1st Author's Affiliation Tokyo University of Agriculture and Technology (TUAT)
2nd Author's Name Shuji Komeiji  
2nd Author's Affiliation Tokyo University of Agriculture and Technology (TUAT)
3rd Author's Name Takumi Mitsuhashi  
3rd Author's Affiliation Juntendo University (Juntendo Univ.)
4th Author's Name Yasushi Iimura  
4th Author's Affiliation Juntendo University (Juntendo Univ.)
5th Author's Name Hiroharu Suzuki  
5th Author's Affiliation Juntendo University (Juntendo Univ.)
6th Author's Name Hidenori Sugano  
6th Author's Affiliation Juntendo University (Juntendo Univ.)
7th Author's Name Koichi Shinoda  
7th Author's Affiliation Tokyo Institute of Technology (TokyoTech)
8th Author's Name Toshihisa Tanaka  
8th Author's Affiliation Tokyo University of Agriculture and Technology (TUAT)
9th Author's Name  
9th Author's Affiliation ()
10th Author's Name  
10th Author's Affiliation ()
11th Author's Name  
11th Author's Affiliation ()
12th Author's Name  
12th Author's Affiliation ()
13th Author's Name  
13th Author's Affiliation ()
14th Author's Name  
14th Author's Affiliation ()
15th Author's Name  
15th Author's Affiliation ()
16th Author's Name  
16th Author's Affiliation ()
17th Author's Name  
17th Author's Affiliation ()
18th Author's Name  
18th Author's Affiliation ()
19th Author's Name  
19th Author's Affiliation ()
20th Author's Name  
20th Author's Affiliation ()
Speaker
Date Time 2021-03-04 16:10:00 
Presentation Time 25 
Registration for SIP 
Paper # EA2020-87, SIP2020-118, SP2020-52 
Volume (vol) 120 
Number (no) no.397(EA), no.398(SIP), no.399(SP) 
Page pp.164-169 
#Pages
Date of Issue 2021-02-24 (EA, SIP, SP) 


[Return to Top Page]

[Return to IEICE Web Page]


The Institute of Electronics, Information and Communication Engineers (IEICE), Japan