• MALSORI
• /
• no.62
• /
• pp.97-112
• /
• 2007

The speech recognizer in general operates in noisy acoustical environments. Many research works have been done to cope with the acoustical variations. Among them, the multiple-HMM model approach seems to be quite effective compared with the conventional methods. In this paper, we consider a multiple-model approach combined with the model compensation method and investigate the necessary number of the HMM model sets through noisy speech recognition experiments. By using the data-driven Jacobian adaptation for the model compensation, the multiple-model approach with only a few model sets for each noise type could achieve comparable results with the re-training method.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.9 no.5
• /
• pp.543-548
• /
• 2014

Multiple-model based speech recognition framework (MMSR) has been known to be very successful in speech recognition. Since it uses multiple hidden Markov modes (HMMs) that corresponds to various noise types and signal-to-noise ratio (SNR) values, the selected acoustic model can have a close match with the test noisy speech. However, since the number of HMM sets is limited in practical use, the acoustic mismatch still remains as a problem. In this study, we experimentally determined the optimal SNR mapping between the test noisy speech and the HMM set to mitigate the mismatch between them. Improved performance was obtained by employing the SNR mapping instead of using the estimated SNR from the test noisy speech. When we applied the proposed method to the MMSR, the experimental results on the Aurora 2 database show that the relative word error rate reduction of 6.3% and 9.4% was achieved compared to a conventional MMSR and multi-condition training (MTR), respectively.

• Phonetics and Speech Sciences
• /
• v.2 no.1
• /
• pp.87-95
• /
• 2010

In this paper, we address issues associated with mixed-lingual text-to-speech based on context-dependent HMMs, where there are multiple sets of HMMs corresponding to each individual language. In particular, we propose smoothing techniques of synthesis parameters at the boundaries between different languages to obtain more natural quality of speech. In other words, mel-frequency cepstral coefficients (MFCCs) at the language boundaries are smoothed by applying several linear and nonlinear approximation techniques. It is shown from an informal listening test that synthesized speech smoothed by a modified version of linear least square approximation (MLLSA) and a quadratic interpolation (QI) method is preferred than that without using any smoothing technique.

• Korean Journal of Microbiology
• /
• v.39 no.4
• /
• pp.207-215
• /
• 2003

Gene structure prediction, which is to predict protein coding regions in a given nucleotide sequence, is the most important process in annotating genes and greatly affects gene analysis and genome annotation. As eukaryotic genes have more complicated stuructures in DNA sequences than those of prokaryotic genes, analysis programs for eukaryotic gene structure prediction have more diverse and more complicated computational models. We have developed EGSP, a eukaryotic gene structure program, using duration hidden markov model. The program consists of two major processes, one of which is a training process to produce parameter values from training data sets and the other of which is to predict protein coding regions based on the parameter values. The program predicts multiple genes rather than a single gene from a DNA sequence. A few computational models were implemented to detect signal pattern and their scanning efficiency was tested. Prediction performance was calculated and was compared with those of a few commonly used programs, GenScan, GeneID and Morgan based on a few criteria. The results show that the program can be practically used as a stand-alone program and a module in a system. For gene prediction of eukaryotic microbial genomes, training and prediction analysis was done with Saccharomyces chromosomes and the result shows the program is currently practically applicable to real eukaryotic microbial genomes.