Title:

Broadcast News: Features and acoustic modelling

Author(s):

Ellis, Daniel P. W.

Date:

1998

Type:

Presentations

Department:

Electrical Engineering

Permanent URL:

http://hdl.handle.net/10022/AC:P:14237

Notes:

Presented to the final review meeting of the SPRACH project at the Institute for Systems and Computer Engineering, Lisbon, Portugal, December 15, 1998.

Abstract:

Overview of the International Computer Science Institute's (ICSI) contributions to a system developed by the European Union-sponsored collaborative Speech Recognition Algorithms for Connectionist Hybrids (SPRACH) project for submission to a Defense Advanced Research Projects Agency and National Institute of Standards and Technology evaluation of approaches to speech recognition of broadcast news footage. ICSI's contributions were in acoustic modelling, in particular modulation-filtered spectrogram features and very large multi-layer perceptron classifiers.

Subject(s):

Electrical engineering
Artificial intelligence

Item views:

27

Metadata:

description

View | Download