Artificial intelligence

Scientific journal

ISSN 2710-1673

ONLINE: ISSN 2710-1681

Search by:

Year of publication
Author name
Paper title

Text-Independent Speaker Recognition Using Another Language Speech Corpora

Sazhok M.1, Seliukh R.1, Fedoryn D.1, Yukhymenko O.1
1 International Research and Training Centre for Information Technologies and Systems

Full text (PDF)

UDC: 004.93
Publication Language: Ukrainian
Stuc. intelekt. 2016; 21(4):37-41

Abstract: This paper presents an experimental research on text-independent speaker recognition using a foreign language speech corpus. Such a technique would allow for involving a large speech corpus when no sufficient speech data available for a given language. Autors considered the open and free speech database THUYG-20 SRE as a foreign corpus. Research results are discussed.

Keywords: text-independent speaker recognition, i-vectors, probabilistic linear discriminant analysis (plda)

References:

  1. Povey D. The Kaldi speech recognition toolkit / Povey D. et all // ASRU, 2011.
  2. Dehak N. Front-End Factor Analysis for Speaker Verification / N. Dehak, P. Kenny, R. Dehak, P. Dumouchel, P. Ouellet // IEEE Trans. Audio, Speech & Language Processing 19(4): 788-798, 2011.
  3. Rozi A. An open/free database and benchmark for uygur speaker recognition / A. Rozi, D. Wang, Z. Zhang // Oriental COCOSDA, 2015.
  4. Cieri C. The Fisher corpus: a resource for the next generations of speech-to-text / C. Cieri, D. Miller, K. Walker // 4th IC on LRE, 2004.
  5. Sazhok M. Adaptatsiya akustychnykh modeley fonem do holosu dyktora dlya pofonemnoho rozpiznavannya izol'ovanykh sliv ukrayins'koyi movy / M. Sazhok, R. Selyukh, O. Yukhymenko // Shtuchnyy intelekt. - 2009. - #4. - S. 230-233.

View full text (PDF)