Layered Nonnegative Matrix Factorization for Speech Separation
顯示簡易詮釋資料 統計資料| 欄位 | 內容 | 語言 |
|---|---|---|
dc.citation.epage |
632 | en_US |
dc.citation.spage |
628 | en_US |
dc.citation.woscount |
0 | en_US |
dc.contributor.author |
Hsu, Chung-Chien | en_US |
dc.contributor.author |
Chien, Jen-Tzung | en_US |
dc.contributor.author |
Chi, Tai-Shih | en_US |
dc.contributor.department |
電機學院 | zh_TW |
dc.contributor.department |
College of Electrical and Computer Engineering | en_US |
dc.date.accessioned |
2017-04-21T06:49:29Z | |
dc.date.available |
2017-04-21T06:49:29Z | |
dc.date.issued |
2015 | en_US |
dc.description.abstract |
This paper proposes a layered nonnegative matrix factorization (L-NMF) algorithm for speech separation. The standard NMF method extracts parts-based bases out of nonnegative training data and is often used to separate mixed spectrograms. The proposed L-NMF algorithm comprises of several layers of standard NMF blocks. During training, each layer of the L-NMF is initialized separately and then fine-tuned by minimizing the propagated reconstruction error. More complicated bases of the training data are emerged in deeper layers of the L-NMF by progressively combining parts-based bases extracted in the first layer. In other words, these complicated bases contain collective information of the parts-based bases. The bases deciphered by all layers are then used to separate spectrograms in the conventional NMF way. Simulation results show the proposed L-NMF outperforms the standard NMF in terms of the source-to-distortion ratio (SDR). | en_US |
dc.identifier.isbn |
978-1-5108-1790-6 | en_US |
dc.identifier.journal |
16TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2015), VOLS 1-5 | en_US |
dc.identifier.uri |
https://ir.lib.nycu.edu.tw/handle/11536/136224 | |
dc.identifier.wosnumber |
WOS:000380581600127 | en_US |
dc.language.iso |
en_US | en_US |
dc.subject |
Layered NMF | en_US |
dc.subject |
dictionary learning | en_US |
dc.subject |
NMF | en_US |
dc.subject |
speech separation | en_US |
dc.title |
Layered Nonnegative Matrix Factorization for Speech Separation | en_US |
dc.type |
Proceedings Paper | en_US |
檔案
本條目沒有附加檔案。
所屬典藏: 會議論文
引用此條目
Hsu, C.-C., Chien, J.-T., & Chi, T.-S. (2015). Layered Nonnegative Matrix Factorization for Speech Separation. 16TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2015), VOLS 1-5, 628–632. https://ir.lib.nycu.edu.tw/handle/11536/136224
引用文字依 APA 第 7 版格式自動產生,使用前請再核對。
分享
永久網址:https://ir.lib.nycu.edu.tw/handle/11536/136224