This paper deals with the CLEAR 2007 evaluation on the detection of acoustic events which happen during seminars.The proposed system first converts an audio sequence in a stream of MFCC features, thena detecting/classifying block identifies an acoustic event with time stamps and assign to it a label among all possible event labels.Identification and classification are based on Hidden Markov Models (HMM).The results, measured in terms of two metrics(accuracy and error rate) are obtained applying the implemented system onthe interactive seminars collected under the CHIL project. Final not very good results highlight the task complexity.
An HMM Based System for Acoustic Event Detection
Zieger, Christian
2007-01-01
Abstract
This paper deals with the CLEAR 2007 evaluation on the detection of acoustic events which happen during seminars.The proposed system first converts an audio sequence in a stream of MFCC features, thena detecting/classifying block identifies an acoustic event with time stamps and assign to it a label among all possible event labels.Identification and classification are based on Hidden Markov Models (HMM).The results, measured in terms of two metrics(accuracy and error rate) are obtained applying the implemented system onthe interactive seminars collected under the CHIL project. Final not very good results highlight the task complexity.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
