Noisy Speech Recognition using Wavelets and FCM
| Autor: | M. A. Anusuya, H. Y. Vani |
|---|---|
| Rok vydání: | 2016 |
| Předmět: |
Fuzzy clustering
Computer science business.industry Speech recognition Feature extraction ComputingMethodologies_IMAGEPROCESSINGANDCOMPUTERVISION Pattern recognition Background noise ComputingMethodologies_PATTERNRECOGNITION Wavelet Computer Science::Sound Robustness (computer science) Recognition system Artificial intelligence Cluster analysis business |
| Zdroj: | Proceedings of the 4th International Conference on Information and Network Security. |
| Popis: | Speech recognition rate decreases, as the background noise increases. Robustness of the system declines in the presence of background noise and it varies for different types of noises. This paper presents the application of wavelet technique in the pre-emphasis phase to increase the robustness of the recognition system in the presence of the noise. It also highlights the importance of applying wavelets and Fuzzy Clustering technique in improving the performance of the speech recognition system. Recognition accuracies are compared and tabulated for various combinations of feature extraction techniques and clustering algorithms for both additive and convolution noises. Performance of all these systems are discussed in terms of recognition accuracies. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|