![]() ![]() ![]() In 1994: proceedings Australian conference on speech science and technology.īrocki, Ł., & Marasek, K. FARSDAT-the speech database of Farsi spoken language. Paper presented at the second workshop on Persian language and computer.īijankhan, M., Sheikhzadegan, J., & Roohani, M. In The 2nd workshop on information technology & its disciplines (WITID 2004) (pp. The sharif speaker-independent large vocabulary speech recognition system. Signal and Data Processing, 13(3), 51–62.īabaali, B., & Sameti, H. A state-of-the-art and efficient framework for Persian Speech Reognition. Sharif University of Technology.īabaAli, B. arXiv preprint arXiv:1001.2267.Īzadi Yazdi, S. In Neural information processing systems (NIPS).Īnusuya, M., & Katti, S. Backpropagation in sequential deep belief networks. Paper presented at the The 9th Iranian electrical engineering conference.Īndrew, G., & Bilmes, J. Shenava 1- Persian continuous speech recognition system. A., Bijankhan, M., Sameti, H., & Sheikh Zadegan, J. Shenava 2- Persian continuous speech recognition software Paper presented at the the first workshop on Persian language and computer, Iran.Īlmas Ganj, F., Seyed Salehi, S. Comparing obtained results with the HMM and Kaldi-DNN indicates that using DBLSTM with features extracted from the DBN increases the accuracy of Persian phoneme recognition.Īlmas Ganj, F. Also, using the bidirectional network increases the accuracy of the model in comparison with the unidirectional network, in both deep and shallow networks. The obtained results show that the use of a deep neural network (DNN) compared to a shallow network improves the results. ![]() In this paper, for the first time, the combination of deep belief network (DBN), for extracting features of speech signals, and Deep Bidirectional Long Short-Term Memory (DBLSTM) with Connectionist Temporal Classification (CTC) output layer is used to create an AM on the Farsdat Persian speech data set. One way to enhance the accuracy of them is by improving the acoustic model (AM). One of the existing challenges is increasing the accuracy and efficiency of these systems. Up to now, various methods are used for Automatic Speech Recognition (ASR), and among which the Hidden Markov Model (HMM) and Artificial Neural Networks (ANNs) are the most important ones. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |