| 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 | 2023-06-23 13:50 Streaming End-to-End speech recognition using a CTC decoder with substituted linguistic information Tatsunari Takagi (TUT), Atsunori Ogawa (NTT), Norihide Kitaoka, Yukoh Wakabayashi (TUT)  SP2023-12 |
|
| Abstract | (in Japanese) | (See Japanese page) |
| (in English) | Speech recognition technology has been employed in various fields due to the enhancement of speech recognition model accuracy. However, when the domain of the data used for training differs from that of the data to be recognized, recognition accuracy declines. To address this issue, several approaches utilizing language models trained on extensive text data have been proposed. Recently, the Density Ratio Approach (DRA), an extension of Shallow Fusion, has been introduced as a method for integrating speech recognition models with language models. Nevertheless, the application of DRA to streaming speech recognition models using a CTC decoder for Japanese speech has not been investigated. In this study, we conducted domain adaptation using DRA in streaming speech recognition with a CTC decoder. To facilitate streaming processing, the decoder successively replaces linguistic information on a frame-by-frame basis, obtaining recognition results through greedy search. Furthermore, we selected the linguistic information to be replaced, considering the assumption of conditional independence of CTC. Experimental results demonstrate that the proposed method enhances recognition accuracy. |
|
| Keyword | (in Japanese) | (See Japanese page) |
| (in English) | End-to-End Speech Recognition / Streaming Speech Recognition / Language Mode / CTC / DRA / / / | |
| Reference Info. | IEICE Tech. Rep., vol. 123, no. 88, SP2023-12, pp. 60-64, June 2023. | |
| Paper # | SP2023-12 | |
| Date of Issue | 2023-06-16 (SP) | |
| ISSN | 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 | SP2023-12 |
|
| Conference Information | |
| Committee | SP IPSJ-MUS IPSJ-SLP |
| Conference Date | 2023-06-23 - 2023-06-24 |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Paper Information | |
| Registration To | SP |
| Conference Code | 2023-06-SP-MUS-SLP |
| Language | Japanese |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Streaming End-to-End speech recognition using a CTC decoder with substituted linguistic information |
| Sub Title (in English) | |
| Keyword(1) | End-to-End Speech Recognition |
| Keyword(2) | Streaming Speech Recognition |
| Keyword(3) | Language Mode |
| Keyword(4) | CTC |
| Keyword(5) | DRA |
| Keyword(6) | |
| Keyword(7) | |
| Keyword(8) | |
| 1st Author's Name | Tatsunari Takagi |
| 1st Author's Affiliation | Toyohashi Univerdity of Technology (TUT) |
| 2nd Author's Name | Atsunori Ogawa |
| 2nd Author's Affiliation | NIPPON TELEGRAPH AND TELEPHONE CORPORATION (NTT) |
| 3rd Author's Name | Norihide Kitaoka |
| 3rd Author's Affiliation | Toyohashi Univerdity of Technology (TUT) |
| 4th Author's Name | Yukoh Wakabayashi |
| 4th Author's Affiliation | Toyohashi Univerdity of Technology (TUT) |
| 5th Author's Name | |
| 5th Author's Affiliation | () |
| 6th Author's Name | |
| 6th Author's Affiliation | () |
| 7th Author's Name | |
| 7th Author's Affiliation | () |
| 8th Author's Name | |
| 8th Author's Affiliation | () |
| 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 | () |
| Speaker | Author-1 |
| Date Time | 2023-06-23 13:50:00 |
| Presentation Time | 140 minutes |
| Registration for | SP |
| Paper # | SP2023-12 |
| Volume (vol) | vol.123 |
| Number (no) | no.88 |
| Page | pp.60-64 |
| #Pages | 5 |
| Date of Issue | 2023-06-16 (SP) |
[Return to Top Page]