![]() Rabiner, L., Juang, B.H.: Fundamentals of Speech Recognition. Huang, X.D., Acero, A., Hon, H.: Spoken language processing. ECTI Transaction on Computer and Information Technology 1(2) (November 2005) In: SPECOM 2006, June 25 (2006)įurui, S.: 50 Years of Progress in Speech and Speaker Recognition Research. Rabiner, L.: Challenges in Speech Recognition and Natural Language Processing. By taking the advantages of the robustness methods, performance of Nevisa in real environments is similar to clean condition. To evaluate the system accuracy, a clean test set was selected from Farsdat and four noisy tasks with different noise types were recorded in different real environments. ![]() For training Nevisa, Farsdat database was used. Some of these methods were modified to achieve higher robustness. Model-based approaches like PMC, MLLR, and MAP, feature robustness methods like CMS, PCA, RCC, and VTLN, and speech enhancement methods like spectral subtraction and Wiener filtering were investigated. In order to compensate the effect of accuracy reduction in noisy environments, powerful robustness methods are utilized. Nevisa is equipped with out-of-vocabulary capability for applications with small size vocabulary. Language modeling for Persian has been implemented in two statistical (n-gram) and grammatical forms. The system is based on phoneme modeling and utilizes synchronous beam search based on lexicon tree for decoding the acoustic utterances. Maximum likelihood estimation criterion the core of which are the classical segmental k-means and Baum-Welsh algorithms is used for training the acoustic models. It also utilizes a VAD based on signal energy and zero-crossing rate. Like most successful recognition systems, MFCC with some modification has been used as speech signal features. Nevisa is an HMM-based, large vocabulary speaker-independent continuous speech recognition system. ![]() ![]() In this paper we have reviewed Nevisa Persian speech recognition engine. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |