Publication detail

Evaluation of Blind Separated Signals Using Speech Recognition System

EKSLER, V.

Original Title

Evaluation of Blind Separated Signals Using Speech Recognition System

Type

conference paper

Language

English

Original Abstract

Output blind separated signals are recovered independent sources obtained from their mixtures using methods of convolutive independent component analysis. These methods have no information about the original sources and the mixing process. For the separation process we adapted and improved a method which is based on the natural gradient approach. The new concept proposed in this paper then tries to measure the hit rate of the source restoration by means of the discrete word intelligibility using an Internet speech recognition system. We are interested in a two-channel case, in which we observe the outputs of a 2 x 2 linear time invariant system of mixtures of speech, music and noise signals.

Keywords

Convolutive ICA, independent component analysis, microphone array, learning algorithm, natural gradient, speech enhancement, speech recognition.

Authors

EKSLER, V.

RIV year

2005

Released

21. 11. 2005

Publisher

EUROCON

Location

Belgrade, Serbia and Montenegro

ISBN

1-4244-0050-3

Book

EUROCON 2005 - The International Conference on "Computer as a tool", R23 Signal Processing, IEEE Catalog Number: 05EX1255C

Pages from

1650

Pages to

1653

Pages count

4

BibTex

@inproceedings{BUT15411,
  author="Václav {Eksler}",
  title="Evaluation of Blind Separated Signals Using Speech Recognition System",
  booktitle="EUROCON 2005 - The International Conference on {"}Computer as a tool{"}, R23 Signal Processing, IEEE Catalog Number: 05EX1255C",
  year="2005",
  pages="4",
  publisher="EUROCON",
  address="Belgrade, Serbia and Montenegro",
  isbn="1-4244-0050-3"
}