Loading...

A fast two-level Speaker Identification method employing sparse representation and GMM-based methods

Zeinali, H ; Sharif University of Technology | 2012

948 Viewed
  1. Type of Document: Article
  2. DOI: 10.1109/ISSPA.2012.6310594
  3. Publisher: 2012
  4. Abstract:
  5. In large population Speaker Identification (SI), computation time has become one of the most important issues in recent real time systems. Test computation time depends on the cost of likelihood computation between test features and registered speaker models. For real time application of speaker identification, system must identify an unknown speaker quickly. Hence the conventional SI methods cannot be used. In this paper, we propose a two-step method that utilizes two different identification methods. In the first step we use Nearest Neighbor method to decrease the search space. In the second step we use GMM-based SI methods to specify the target speaker. We achieved 3.5× speed-ups without any loss of accuracy using the proposed method. If the number of best speaker is reduced, the Identification accuracy decreases. So, there is a trade-off between accuracy and speed-up
  6. Keywords:
  7. Computation time ; Identification accuracy ; Identification method ; Large population ; Likelihood computation ; Loss of accuracy ; Nearest neighbor method ; Real-time application ; Search spaces ; Sparse representation ; Speaker identification ; Speaker model ; Target speaker ; Two step method ; Information science ; Interactive computer systems ; Real time systems ; Signal processing ; Loudspeakers
  8. Source: 2012 11th International Conference on Information Science, Signal Processing and their Applications, ISSPA 2012 ; 2012 , Pages 45-48 ; 9781467303828 (ISBN)
  9. URL: http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6310594