Chapter 1: Invited Addresses and Tutorials on Signals, Coding,
  Systems and Intelligent Techniques

Title: Speaker Identification Using Orthogonal and Discriminative Features

Author(s): Seyyed Hashem Davarpanah, Abdolreza Mirzaei, Amir Ziaei

Address: Department of Computer Engineering, Islamic Azad University Malayer Branch, Malayer County, Iran | Department of Computer Engineering, Amir Kabir University of Technology, Valiasr, Tehran, Iran | Department of Computer Engineering, Shamsipour Inistitue, Vanak, Tehran, Iran

Reference: 12th International Workshop on Systems, Signals and Image Processing pp. 295 - 298

Abstract/Summary: It is desirable that acoustic vectors would form separable clusters in the feature space; however analysis of the common feature vectors does not support this assumption. This paper proposes a new method that manipulates the original features to produce a new feature set in which classes have more convex shape. The proposed methodology uses the idea that according to it, different features have unequal discrimination properties between speakers; So an automatic weighting function based on Multivariate Analysis Of Variance Algorithm (MANOVA) is proposed. MANOVA searches for a linear combination of original features with the largest separation among the speakers. The Vector Quantization (VQ) algorithm is used to detect speakers in the next stage. Although this algorithm is faster and has fewer complexes than the other classification algorithms in this content, promising results are achieved.

Order a copy of this article