Conference paper
Audio-emotion recognition system using parallel classifiers and audio feature analyzer
Proceedings of the 3rd International Conference on Computational Intelligence, Modelling and Simulation, pp.210-215
International Conference on Computational Intelligence, Modelling & Simulation, 3rd (Langkawi, Malaysia, 20-Sep-2011–22-Sep-2011)
Institute of Electrical and Electronics Engineers
2011
Abstract
Emotion recognition based on an audio signal is an area of active research in the domain of human-computer interaction and effective computing. This paper presents an audio-emotion recognition (AER) system using parallel classifiers and an audio feature analyzer. In the proposed system, audio features such as the pitch and fractional cepstral coefficient are first extracted from the audio signal for analysis. These extracted features are then used to train a radial basis function. Lastly, an audio feature analyzer is used to enhance the performance of the recognition rate. The latest simulation results show that the proposed AER system is able to achieve an emotion recognition rate of 81.67%.
Details
- Title
- Audio-emotion recognition system using parallel classifiers and audio feature analyzer
- Authors
- Li Wern Chew (Author) - University of Nottingham Malaysia CampusK P Seng (Author) - Sunway UniversityLi-Minn Ang (Author) - University of Nottingham Malaysia CampusVish Ramakonar (Author) - Alsys MSC Sdn Bhd (Kuala Lumpur, Malaysia)A Gnanasegaran (Author) - Alsys MSC Sdn Bhd (Kuala Lumpur, Malaysia)
- Publication details
- Proceedings of the 3rd International Conference on Computational Intelligence, Modelling and Simulation, pp.210-215
- Conference details
- International Conference on Computational Intelligence, Modelling & Simulation, 3rd (Langkawi, Malaysia, 20-Sep-2011–22-Sep-2011)
- Publisher
- Institute of Electrical and Electronics Engineers
- Date published
- 2011
- DOI
- 10.1109/CIMSim.2011.44
- ISSN
- 2166-8531; 2166-8523
- Organisation Unit
- University of the Sunshine Coast, Queensland; School of Science, Technology and Engineering; Engage Research Lab
- Language
- English
- Record Identifier
- 99513896302621
- Output Type
- Conference paper
Metrics
78 Record Views