完整後設資料紀錄
DC 欄位語言
dc.contributor.authorHsu, Chung-Chienen_US
dc.contributor.authorChien, Jen-Tzungen_US
dc.contributor.authorChi, Tai-Shihen_US
dc.date.accessioned2017-04-21T06:49:29Z-
dc.date.available2017-04-21T06:49:29Z-
dc.date.issued2015en_US
dc.identifier.isbn978-1-5108-1790-6en_US
dc.identifier.urihttp://hdl.handle.net/11536/136224-
dc.description.abstractThis 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.language.isoen_USen_US
dc.subjectLayered NMFen_US
dc.subjectdictionary learningen_US
dc.subjectNMFen_US
dc.subjectspeech separationen_US
dc.titleLayered Nonnegative Matrix Factorization for Speech Separationen_US
dc.typeProceedings Paperen_US
dc.identifier.journal16TH ANNUAL CONFERENCE OF THE INTERNATIONAL SPEECH COMMUNICATION ASSOCIATION (INTERSPEECH 2015), VOLS 1-5en_US
dc.citation.spage628en_US
dc.citation.epage632en_US
dc.contributor.department電機學院zh_TW
dc.contributor.departmentCollege of Electrical and Computer Engineeringen_US
dc.identifier.wosnumberWOS:000380581600127en_US
dc.citation.woscount0en_US
顯示於類別:會議論文