• MALSORI
• /
• no.62
• /
• pp.133-147
• /
• 2007

In this paper, we introduce a phone vector discrete HMM(PVDHMM) that decodes a phone sequence string, and demonstrates the applicability to spoken document retrieval. The PVDHMM treats a phone recognizer or large vocabulary continuous speech recognizer (LVCSR) as a vector quantizer whose codebook size is equal to the size of its phone set. We apply the PVDHMM to decode the phone sequence strings and compare the outputs with those of a continuous speech recognizer(CSR). Also we carry out spoken document retrieval experiment through PVDHMM word spotter on the phone sequence strings which are generated by phone recognizer or LVCSR and compare its results with those of retrieval through the phone-based vector space model.

• Phonetics and Speech Sciences
• /
• v.7 no.2
• /
• pp.149-156
• /
• 2015

We propose a new method for automatic fluency scoring of English speaking tests spoken by nonnative speakers in a free-talking style. The proposed method is different from the previous methods in that it does not require the transcribed texts for spoken utterances. At first, an input utterance is segmented into a phone sequence by using a phone recognizer trained by using native speech databases. For each utterance, a feature vector with 6 features is extracted by processing the segmentation results of the phone recognizer. Then, fluency score is computed by applying support vector regression (SVR) to the feature vector. The parameters of SVR are learned by using the rater scores for the utterances. In computer experiments with 3 tests taken by 48 Korean adults, we show that speech rate, phonation time ratio, and smoothed unfilled pause rate are best for fluency scoring. The correlation of between the rater score and the SVR score is shown to be 0.84, which is higher than the correlation of 0.78 among raters. Although the correlation is slightly lower than the correlation of 0.90 when the transcribed texts are given, it implies that the proposed method can be used as a preprocessing tool for fluency evaluation of speaking tests.

• Proceedings of the KSPS conference
• /
• 2007.05a
• /
• pp.131-134
• /
• 2007

We propose a multi-stage recognizer architecture that reduces the computation load and makes fast recognizer. To improve performance of baseline multi-stage recognizer, we introduced new feature. We used confidence vector for each phone segment instead of best phoneme sequence. The multi-stage recognizer with new feature has better performance on n-best and has more robustness.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1998.08a
• /
• pp.87-92
• /
• 1998

In general, the synthesis unit database was constructed by recording isolated word. In that case, each boundary of word has typical prosodic pattern like a falling intonation or preboundary lengthening. To get natural synthetic speech using these kinds of database, we must artificially distort original speech. However, that artificial process rather resulted in unnatural, unintelligible synthetic speech due to the excessive prosodic modification on speech signal. To overcome these problems, we gathered thousands of sentences for synthesis database. To make a phone level synthesis unit, we trained speech recognizer with the recorded speech, and then segmented phone boundaries automatically. In addition, we used laryngo graph for the epoch detection. From the automatically generated synthesis database, we chose the best phone and directly concatenated it without any prosody processing. To select the best phone among multiple phone candidates, we used prosodic information such as break strength of word boundaries, phonetic contexts, cepstrum, pitch, energy, and phone duration. From the pilot test, we obtained some positive results.

• The Journal of the Acoustical Society of Korea
• /
• v.22 no.3
• /
• pp.194-201
• /
• 2003

In speech recognition, confidence measuring is to decide whether it can be accepted as the recognized results or not. The confidence is measured by integrating frames into phone and word level. In case of word recognition, the confidence measuring verifies the results of recognition and Out-Of-Vocabulary (OOV). Therefore, the post-processing could improve the performance of recognizer without accepting it as a recognition error. In this paper, we measure the confidence modifying log likelihood ratio (LLR) which was the previous confidence measuring. It accumulates only those which the log likelihood ratio is negative when integrating the confidence to phone level from frame level. When comparing the verification performance for the results of word recognizer with the previous method, the FAR (False Acceptance Ratio) is decreased about 3.49% for the OOV and 15.25% for the recognition error when CAR (Correct Acceptance Ratio) is about 90%.

• The Journal of the Acoustical Society of Korea
• /
• v.25 no.1E
• /
• pp.20-24
• /
• 2006

In this paper, we focused on the real time implementation of a speech recognition system with medium size of vocabulary considering its application to a mobile phone. First, we developed the PC based variable vocabulary word recognizer having the size of program memory and total acoustic models as small as possible. To reduce the memory size of acoustic models, linear discriminant analysis and phonetic tied mixture were applied in the feature selection process and training HMMs, respectively. In addition, state based Gaussian selection method with the real time cepstral normalization was used for reduction of computational load and robust recognition. Then, we verified the real-time operation of the implemented recognition system on the TMS320C6201 EVM board. The implemented recognition system uses memory size of about 610 kbytes including both program memory and data memory. The recognition rate was 95.86% for ETRI 445DB, and 96.4%, 97.92%, 87.04% for three kinds of name databases collected through the mobile phones.

• The Journal of Information Technology
• /
• v.3 no.4
• /
• pp.1-12
• /
• 2000

We can devide the speech into high frequency speech and low frequency speech according to the feature of the speech, However so far the construction of the recognizer without concerning this feature causes low recognition rate relatively and the needs of an amount of data in the research on the speech recognition. In this paper, we propose the method that can devide this feature of speaker's speech using the Formant frequency, and the method that can recognize the speech after constructing the recognizer model reflecting the feature of the high and low frequency of the speaker's speech, For the experiment we constructed the recognizer model using 47 mono-phone of Korean and trained the recognizer model using 20 women's and men's speech respectively. We divided the feature of speech using the Formant frequency Table, that had been consisted of the Formant frequency, and the value of pitch, and then We performed recognition using the trained model according to the feature of speech The proposed system outperformed the existing method in the recognition rate, as the result.

• The Journal of the Acoustical Society of Korea
• /
• v.16 no.2
• /
• pp.31-35
• /
• 1997

In this paper, we propose a variable vocabulary word recognizer that is able to recognize new words not exist in training data. For the variable vocabulary word recognizer, we must have an on-line lexicon generator to transform new candidate words to the corresponding pronunciation sequences of phones without any large lexicon table. And, we also must make outputs. In order to model the phones and allophones reliably, we define Korean allophones by triphone clustering based on phonetic knowledge of preceding and succeeding phones of each phone. Using the clustering method, we generated 1,548 allophones with POW (Phonetically Optimized Words) 3,848 word DB. We evaluated the proposed word recognizer with POW 3,848 DB, PBW (Phonetically Balanced Words) 445 DB, and 244 word DB in hotel reservation task. Experimental results showed word recognition accuracy of 79.6% for the POW DB corresponding to vocabulary-dependent case, 79.4% in case of 445 word lexicon and 88.9% in case of 100 word lexicon for the PBW DB, and 71.4% for the hotel reservation DB corresponding to vocabulary-independent case.

• The Journal of the Acoustical Society of Korea
• /
• v.18 no.3
• /
• pp.3-9
• /
• 1999

This paper presents the study on the rejection capability based on anti-phone modeling for vocabulary independent speech recognition system. The rejection system detects and rejects out-of-vocabulary words which were not included in candidate words which are defined while the speech recognizer is made. The rejection system can be classified into two categories by their implementation methods, keyword spotting method and utterance verification method. The keyword spotting method uses an extra filler model as a candidate word as well as keyword models. The utterance verification method uses the anti-models for each phoneme for the calculation of confidence score after it has constructed the anti-models for all phonemes. We implemented an utterance verification algorithm which can be used for vocabulary independent speech recognizer. We also compared three kinds of means for the calculation of confidence score, and found out that the geometric mean had shown the best result. For the normalization of confidence score, usually Sigmoid function is used. On using it, we compared the effect of the weight constant for Sigmoid function and determined the optimal value. And we compared the effects of the size of cohort set, the results showed that the larger set gave the better results. And finally we found out optimal confidence score threshold value. In case of using the threshold value, the overall recognition rate including rejection errors was about 76%. This results are going to be adapted for stock information system based on speech recognizer which is currently provided as an experimental service by Korea Telecom.

• MALSORI
• /
• no.67
• /
• pp.149-166
• /
• 2008

In spoken document retrievals (SDR), subword (typically phonemes) indexing term is used to avoid the out-of-vocabulary (OOV) problem. It makes the indexing and retrieval process independent from any vocabulary. It also requires a small corpus to train the acoustic model. However, subword indexing term approach has a major drawback. It shows higher word error rates than the large vocabulary continuous speech recognition (LVCSR) system. In this paper, we propose an probabilistic slot detection and n-gram based string matching method for phone based spoken document retrievals to overcome high error rates of phone recognizer. Experimental results have shown 9.25% relative improvement in the mean average precision (mAP) with 1.7 times speed up in comparison with the baseline system.