| Paper Abstract and Keywords |
| Presentation |
2024-03-01 16:05
Evaluating speech generation based on objective measures for text generation Takaaki Saeki (UTokyo), Soumi Maiti (CMU), Shinnosuke Takamichi (UTokyo), Shinji Watanabe (CMU), Hiroshi Saruwatari (UTokyo) EA2023-133 SIP2023-180 SP2023-115 |
| Abstract |
(in Japanese) |
(See Japanese page) |
| (in English) |
In the evaluation of speech generation, while subjective judgments have long been the gold standard, objective metrics such as Mel Cepstral Distortion (MCD) and Mean Opinion Score (MOS) prediction models have also been used. These objective metrics are valued for their lower time and financial costs and the ability to compare different results, driving the demand for metrics that correlate well with human subjective judgments. This paper proposes an automatic evaluation method for speech generation based on text generation metrics. Our proposed SpeechBERTScore computes BERTScores on self-supervised speech feature sequences derived from both generated and reference speech. In addition, SpeechBLEU and SpeechTokenDistance define metrics using self-supervised discrete speech tokens. Experimental evaluations of synthesized speech show that our SpeechBERTScore correlates better with human subjective ratings than MCD and state-of-the-art MOS prediction models. Furthermore, our approach is effective for degraded speech evaluation and applicable across languages. |
| Keyword |
(in Japanese) |
(See Japanese page) |
| (in English) |
Automatic speech evaluation / Speech generation / Self-supervised speech representations / Text generation metrics / / / / |
| Reference Info. |
IEICE Tech. Rep., vol. 123, no. 403, SP2023-115, pp. 421-426, Feb. 2024. |
| Paper # |
SP2023-115 |
| Date of Issue |
2024-02-22 (EA, SIP, 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 |
EA2023-133 SIP2023-180 SP2023-115 |
| Conference Information |
| Committee |
SIP SP EA IPSJ-SLP |
| Conference Date |
2024-02-29 - 2024-03-01 |
| 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 |
2024-02-SIP-SP-EA-SLP |
| Language |
Japanese |
| Title (in Japanese) |
(See Japanese page) |
| Sub Title (in Japanese) |
(See Japanese page) |
| Title (in English) |
Evaluating speech generation based on objective measures for text generation |
| Sub Title (in English) |
|
| Keyword(1) |
Automatic speech evaluation |
| Keyword(2) |
Speech generation |
| Keyword(3) |
Self-supervised speech representations |
| Keyword(4) |
Text generation metrics |
| Keyword(5) |
|
| Keyword(6) |
|
| Keyword(7) |
|
| Keyword(8) |
|
| 1st Author's Name |
Takaaki Saeki |
| 1st Author's Affiliation |
The University of Tokyo (UTokyo) |
| 2nd Author's Name |
Soumi Maiti |
| 2nd Author's Affiliation |
Carnegie Mellon University (CMU) |
| 3rd Author's Name |
Shinnosuke Takamichi |
| 3rd Author's Affiliation |
The University of Tokyo (UTokyo) |
| 4th Author's Name |
Shinji Watanabe |
| 4th Author's Affiliation |
Carnegie Mellon University (CMU) |
| 5th Author's Name |
Hiroshi Saruwatari |
| 5th Author's Affiliation |
The University of Tokyo (UTokyo) |
| 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 |
() |
| 19th Author's Name |
|
| 19th Author's Affiliation |
() |
| 20th Author's Name |
|
| 20th Author's Affiliation |
() |
| 21st Author's Name |
|
| 21st Author's Affiliation |
() |
| 22nd Author's Name |
|
| 22nd Author's Affiliation |
() |
| 23rd Author's Name |
|
| 23rd Author's Affiliation |
() |
| 24th Author's Name |
|
| 24th Author's Affiliation |
() |
| 25th Author's Name |
|
| 25th Author's Affiliation |
() |
| 26th Author's Name |
/ / |
| 26th Author's Affiliation |
()
() |
| 27th Author's Name |
/ / |
| 27th Author's Affiliation |
()
() |
| 28th Author's Name |
/ / |
| 28th Author's Affiliation |
()
() |
| 29th Author's Name |
/ / |
| 29th Author's Affiliation |
()
() |
| 30th Author's Name |
/ / |
| 30th Author's Affiliation |
()
() |
| 31st Author's Name |
/ / |
| 31st Author's Affiliation |
()
() |
| 32nd Author's Name |
/ / |
| 32nd Author's Affiliation |
()
() |
| 33rd Author's Name |
/ / |
| 33rd Author's Affiliation |
()
() |
| 34th Author's Name |
/ / |
| 34th Author's Affiliation |
()
() |
| 35th Author's Name |
/ / |
| 35th Author's Affiliation |
()
() |
| 36th Author's Name |
/ / |
| 36th Author's Affiliation |
()
() |
| Speaker |
Author-1 |
| Date Time |
2024-03-01 16:05:00 |
| Presentation Time |
20 minutes |
| Registration for |
SP |
| Paper # |
EA2023-133, SIP2023-180, SP2023-115 |
| Volume (vol) |
vol.123 |
| Number (no) |
no.401(EA), no.402(SIP), no.403(SP) |
| Page |
pp.421-426 |
| #Pages |
6 |
| Date of Issue |
2024-02-22 (EA, SIP, SP) |