Beschreibung
Our proposed method detects emotions from speech using prosodic and spectral features. It uses pitch and MFCC features of speech. This method performs the classification using Artificial Neural Network and Naïve Bayes Classifier. In this work the popular speech database Berlin Emo-Db is used for the classification. This paper also shows how the performance of speech emotion recognition system depends on the number of emotions to detect and the speaker. After performing the classification the accuracy of the system is also tested using the speech samples of Berlin Emo-db and the results are considerably better than many existing systems.
Autorenporträt
Atreyee Khan has completed M. E. in Software Engineering from Jadavpur University. She has expertise in Java and Android domain. Currently, she is working in a multinational bank in analytics domain. Her research interests include voice-based emotion detection, machine learning technique using SAS and R.
Herstellerkennzeichnung:
BoD - Books on Demand
In de Tarpen 42
22848 Norderstedt
DE
E-Mail: info@bod.de
