Speech dimensionality analysis on hypercubical self-organizing maps

被引：6

作者：

Somervuo, P ^{[1
]}

机构：

[1] Aalto Univ, Neural Networks Res Ctr, FIN-02015 Espoo, Finland

来源：

NEURAL PROCESSING LETTERS | 2003年 / 17卷 / 02期

关键词：

dimension reduction; fractal dimension; multi-dimensional scaling; self-organizing map; sequence projection; speech analysis;

D O I：

10.1023/A:1023646203167

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

The problem of finding the intrinsic dimension of speech is addressed in this paper. A structured vector quantization lattice, Self-Organizing Map (SOM), is used as a projection space for the data. The goal is to find a hypercubical SOM lattice where the sequences of projected speech feature vectors form continuous trajectories. The effect of varying the dimension of the lattice is investigated using feature vector sequences computed from the TIMIT database.

引用

页码：125 / 136

页数：12

共 22 条

[1]

Aho A. V., 1983, DATA STRUCTURES ALGO

[2] QUANTIFYING THE NEIGHBORHOOD PRESERVATION OF SELF-ORGANIZING FEATURE MAPS [J].

BAUER, HU ;

PAWELZIK, KR .

IEEE TRANSACTIONS ON NEURAL NETWORKS, 1992, 3 (04) :570-579

[3] Growing a hypercubical output space in a self-organizing feature map [J].

Bauer, HU ;

Villmann, T .

IEEE TRANSACTIONS ON NEURAL NETWORKS, 1997, 8 (02) :218-226

[4]

BISHOP C, 1997, 5 AST U NEUR COMP RE

[5] GTM: The generative topographic mapping [J].

Bishop, CM ;

Svensen, M ;

Williams, CKI .

NEURAL COMPUTATION, 1998, 10 (01) :215-234

[6] Intrinsic dimension estimation of data: An approach based on Grassberger-Procaccia's algorithm [J].

Camastra, F ;

Vinciarelli, A .

NEURAL PROCESSING LETTERS, 2001, 14 (01) :27-34

[7] Curvilinear component analysis: A self-organizing neural network for nonlinear mapping of data sets [J].

Demartines, P ;

Herault, J .

IEEE TRANSACTIONS ON NEURAL NETWORKS, 1997, 8 (01) :148-154

[8]

Frisone F., 1995, P INT C ART NEUR NET, P323

[9] MEASURING THE STRANGENESS OF STRANGE ATTRACTORS [J].

GRASSBERGER, P ;

PROCACCIA, I .

PHYSICA D, 1983, 9 (1-2) :189-208

[10] REPLICATOR NEURAL NETWORKS FOR UNIVERSAL OPTIMAL SOURCE-CODING [J].

HECHTNIELSEN, R .

SCIENCE, 1995, 269 (5232) :1860-1863

← 1 2 3 →