Croatian Large Vocabulary Automatic Speech Recognition

Martinčić-Ipšić, Sanda; Pobar, Miran; Ipšić, Ivo
December 2011
Automatika: Journal for Control, Measurement, Electronics, Compu;2011, Vol. 52 Issue 2, p147
Academic Journal
This paper presents procedures used for development of a Croatian large vocabulary automatic speech recognition system (LVASR). The proposed acoustic model is based on context-dependent triphone hidden Markov models and Croatian phonetic rules. Different acoustic and language models, developed using a large collection of Croatian speech, are discussed and compared. The paper proposes the best feature vectors and acoustic modeling procedures using which lowest word error rates for Croatian speech are achieved. In addition, Croatian language modeling procedures are evaluated and adopted for speaker independent spontaneous speech recognition. Presented experiments and results show that the proposed approach for automatic speech recognition using context-dependent acoustic modeling based on Croatian phonetic rules and a parameter tying procedure can be used for efficient Croatian large vocabulary speech recognition with word error rates below 5%.


Related Articles

  • Isolated Words Recognition System Based on Hybrid Approach DTW/GHMM. Bourouba, E-Hocine; Bedda, Mouldi; Djemili, Rafik // Informatica (03505596);Oct2006, Vol. 30 Issue 3, p373 

    In this paper, we present a new hybrid approach for isolated spoken word recognition using Hidden Markov Model models (HMM) combined with Dynamic time warping (DTW). HMM have been shown to be robust in spoken recognition systems. We propose to extend the HMM method by combining it with the DTW...

  • SPEECH RECOGNITION USING HIDDEN MARKOV MODELS. Sajjan, Sharada C.; Vijaya, C. // World Journal of Science & Technology;2011, Vol. 1 Issue 12, p75 

    This study proposes limited vocabulary isolated word recognition system adopting the Hidden Markov Model to statistically model the words in the dictionary. Feature extraction using Linear Predictive Analysis is carried over the speech frame of 300 samples with 100 samples overlap at 8KHz...

  • Hidden Markov Model for Speech Recognition Using Modified Forward-Backward Re-estimation Algorithm. Sonkamble, Balwant A.; Doye, D. D. // International Journal of Computer Science Issues (IJCSI);Jul2012, Vol. 9 Issue 4, p242 

    There are various kinds of practical implementation issues for the HMM. The use of scaling factor is the main issue in HMM implementation. The scaling factor is used for obtaining smoothened probabilities. The proposed technique called Modified Forward-Backward Re-estimation algorithm used to...

  • Automatic Speech Recognition Technique for Bangla Words. Ali, Md. Akkas; Hossain, Manwar; Bhuiyan, Mohammad Nuruzzaman // International Journal of Advanced Science & Technology;Jan2013, Vol. 50, p51 

    Automatic recognition of spoken words is one of the most challenging tasks in the field of speech recognition. The difficulty of this task is due to the acoustic similarity of many of the words and their syllabi. Accurate recognition requires the system to perform fine phonetic distinctions....

  • A Comparison of DHMM and DTW for Isolated Digits Recognition System of Arabic Language. Hachkar, Z.; Farchi, A.; Mounir, B.; El Abbadi, J. // International Journal on Computer Science & Engineering;2011, Vol. 3 Issue 3, p1002 

    Despite many years of concentrated research, the performance gap between automatic speech recognition (ASR) and human speech recognition (HSR) remains large. Especially for Arabic language, research efforts are still limited in comparison with other languages such as English or Japanese. In this...

  • Reducing computational load in segmental hidden Markov model decoding for speech recognition. Russell, M. J. // Electronics Letters;12/8/2005, Vol. 41 Issue 25, p1408 

    Segment models have the potential to improve automatic speech recognition accuracy but with increased computational load. Two techniques which reduce this load are described: segmental beam pruning, and duration pruning. Experiments show that they can combine to give a 95% reduction in segment...

  • Decision rule based on ordered-nearest-neighbour with applications to utterance verification. Huang, C.-S.; Lee, C.-H.; Wang, H.-C. // Electronics Letters;2/6/2003, Vol. 39 Issue 3, p327 

    Proposes a novel decision rule based on ordered-nearest-neighbor for utterance verification in automatic speech recognition technology. Exploitation of the underlying neighborhood distribution associated with each class as an auxiliary criterion for the plug-in maximum a posteriori decision...

  • EXTENSION OF HIDDEN MARKOV MODEL FOR RECOGNIZING LARGE VOCABULARY OF SIGN LANGUAGE. Jebali, Maher; Dalle, Patrice; Jemni, Mohamed // International Journal of Artificial Intelligence & Applications;Mar2013, Vol. 4 Issue 2, p35 

    Computers still have a long way to go before they can interact with users in a truly natural fashion. From a user's perspective, the most natural way to interact with a computer would be through a speech and gesture interface. Although speech recognition has made significant advances in the past...

  • Arabic Speaker-Independent Continuous Automatic Speech Recognition Based on a Phonetically Rich and Balanced Speech Corpus. Abushariah, Mohammad; Ainon, Raja; Zainuddin, Roziati; Elshafei, Moustafa; Khalifa, Othman // International Arab Journal of Information Technology (IAJIT);Jan2012, Vol. 9 Issue 1, p84 

    This paper describes and proposes an efficient and effective framework for the design and development of a speaker-independent continuous automatic Arabic speech recognition system based on a phonetically rich and balanced speech corpus. The speech corpus contains a total of 415 sentences...


Read the Article


Sorry, but this item is not currently available from your library.

Try another library?
Sign out of this library

Other Topics