WEKO3
アイテム
An Efficient Framework for Constructing Speech Emotion Corpus Based on Integrated Active Learning Strategies
https://tokushima-u.repo.nii.ac.jp/records/2011019
https://tokushima-u.repo.nii.ac.jp/records/201101994e943c6-d229-4bdb-8c55-6b4c18b56147
| 名前 / ファイル | ライセンス | アクション |
|---|---|---|
taffc_13_4_1929.pdf (9.06 MB)
|
|
| Item type | 文献 / Documents(1) | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 公開日 | 2023-07-21 | |||||||||||||
| アクセス権 | ||||||||||||||
| アクセス権 | open access | |||||||||||||
| 資源タイプ | ||||||||||||||
| 資源タイプ識別子 | http://purl.org/coar/resource_type/c_6501 | |||||||||||||
| 資源タイプ | journal article | |||||||||||||
| 出版社版DOI | ||||||||||||||
| 関連識別子 | https://doi.org/10.1109/TAFFC.2022.3192899 | |||||||||||||
| 関連名称 | 10.1109/TAFFC.2022.3192899 | |||||||||||||
| 出版タイプ | ||||||||||||||
| 出版タイプ | AM | |||||||||||||
| 出版タイプResource | http://purl.org/coar/version/c_ab4af688f83e57aa | |||||||||||||
| タイトル | ||||||||||||||
| タイトル | An Efficient Framework for Constructing Speech Emotion Corpus Based on Integrated Active Learning Strategies | |||||||||||||
| 著者 |
任, 福継
× 任, 福継
WEKO
401
× Liu, Zheng
× 康, 鑫 |
|||||||||||||
| 抄録 | ||||||||||||||
| 内容記述 | Speech emotion recognition has been developed rapidly in recent decades because of the appearance of machine learning. Nevertheless, lack of corpus remains a significant issue. For actual speech emotion corpus construction, many professional actors are required to perform voices with various emotions in specific scenes. In the process of data labelling, since the number of samples of different emotion categories is extremely imbalanced, it is difficult to efficiently label the samples. Hence, we proposed an integrated active learning sampling strategy and designed an efficient framework for constructing speech emotion corpora in order to address the problems presented above. Comparing experiments with other active learning algorithms on 13 datasets, our method was shown to improve sampling efficiency. In addition, it is able to select small category samples to be labelled with preference in imbalanced datasets. During the actual corpus construction experiments, our method can prioritize selecting small class emotion samples. As even when the amount of labelled data is less than 50%, the accuracy rate still can reach 90%. This greatly enhances the efficiency of constructing the speech emotion corpus and fills in the gaps. | |||||||||||||
| キーワード | ||||||||||||||
| 主題 | Affective computing | |||||||||||||
| キーワード | ||||||||||||||
| 主題 | Speech emotion recognition | |||||||||||||
| キーワード | ||||||||||||||
| 主題 | Corpus construction | |||||||||||||
| キーワード | ||||||||||||||
| 主題 | Active learning | |||||||||||||
| キーワード | ||||||||||||||
| 主題 | Imbalanced dataset | |||||||||||||
| 書誌情報 |
en : IEEE Transactions on Affective Computing 巻 13, 号 4, p. 1929-1940, 発行日 2022-08-08 |
|||||||||||||
| 収録物ID | ||||||||||||||
| 収録物識別子タイプ | ISSN | |||||||||||||
| 収録物識別子 | 19493045 | |||||||||||||
| 出版者 | ||||||||||||||
| 出版者 | IEEE | |||||||||||||
| 権利情報 | ||||||||||||||
| 権利情報 | © 2022 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. | |||||||||||||
| EID | ||||||||||||||
| 識別子 | 386481 | |||||||||||||
| 言語 | ||||||||||||||
| 言語 | eng | |||||||||||||
