Summary of the paper
| Title | An HMM System for Recognizing Articulation Features for Arabic Phones |
|---|---|
| Authors | Hosam Hammady, Sherif Abdou, Mostafa Shahin, Mohsen Rashwan and Ossama Badawy |
| Abstract | In this paper, we introduce a Hidden Markov Model (HMM) recognition system for the articulation features of Arabic phones. The low-level features are described by Mel- Frequency Cepstral Coefficients (MFCCs). The created HMMs directly model certain articulation features. Articulation features are either place or manner features, here 10 basic manner features are used and arranged in pairs (Adhesion/Separation - Elevation/Lowering - Fluency/Desisting - Plosiveness/Fricativeness - Voicing/Unvoicing). Classification is done on these features regardless of the phone itself. The model has been created successfully and tested on reference speech data. The error rate is very low for many phones and acceptable for most of them. Finally, the system output is used as a confidence measure applied to other existing speech recognizers. |
| Topics | Exploitation of LRs in different types of applications (information extraction, information retrieval, speech dictation, translation, summarisation, web services, semantic web, etc.), LRs for linguistic research in human-machine communication, National and international activities and projects on Arabic |
Full paper  |
An HMM System for Recognizing Articulation Features for Arabic Phones |
| Bibtex | @InProceedings{HAMMADY09.53, author = {Hosam Hammady, Sherif Abdou, Mostafa Shahin, Mohsen Rashwan and Ossama Badawy}, title = {An HMM System for Recognizing Articulation Features for Arabic Phones}, booktitle = {Proceedings of the Second International Conference on Arabic Language Resources and Tools}, year = {2009}, month = {April}, date = {22-23}, address = {Cairo, Egypt}, editor = {Khalid Choukri and Bente Maegaard}, publisher = {The MEDAR Consortium}, isbn = {2-9517408-5-9}, language = {english} } |