| Paper: | MLSP-P4.7 | ||
| Session: | Audio and Communication Applications | ||
| Time: | Thursday, May 18, 14:00 - 16:00 | ||
| Presentation: | Poster | ||
| Topic: | Machine Learning for Signal Processing: Speech and Audio Processing Applications | ||
| Title: | A STUDY OF PERCEPTRON MAPPING CAPABILITY TO DESIGN SPEECH EVENT DETECTORS | ||
| Authors: | Sabato M. Siniscalchi, Mark A. Clements, Georgia Institute of Technology, United States; Antonio Gentile, Giorgio Vassallo, Filippo Sorbello, Università degli Studi di Palermo, Italy | ||
| Abstract: | Event detection is a fundamental yet critical component in automatic speech recognition (ASR) systems that attempt to extract knowledge-based features at the front-end level. In this context, it is common practice to design the detectors inside well-known frameworks based on artificial neural network (ANN) or support vector machine (SVM). In the case of ANN, speech scientists often design their detector architecture relying on conventional feed-forward multi-layer perceptron (MLP) with sigmoidal activation function. The aim of this paper is to introduce other ANN architectures inside the context of detection-based ASR. In particular, a bank of feed-forward MLPs using sinusoidal activation functions is set up to address the event detection problem. Experimental results demonstrate the effectiveness of this ANN design for speech attribute detectors. | ||
©2018 Conference Management Services, Inc. -||- email: webmaster@icassp2006.org -||- Last updated Friday, August 17, 2012