• The Journal of the Acoustical Society of Korea
• /
• v.14 no.5
• /
• pp.5-10
• /
• 1995

In this paper we present a clustering procedure based on the use of HMM in order to get multiple statistical models which can well absorb the variants of each speaker with different ways of saying words. The HMM-clustered models obtained from the developed technique are applied to the speaker independent isolated word recognition. The HMM clustering method splits off all observation sequences with poor likelihood scores which fall below threshold from the training set and create a new model out of the observation sequences in the new cluster. Clustering is iterated by classifying each observation sequence as belonging to the cluster whose model has the maximum likelihood score. If any clutter has changed from the previous iteration the model in that cluster is reestimated by using the Baum-Welch reestimation procedure. Therefore, this method is more efficient than the conventional template-based clustering technique due to the integration capability of the clustering procedure and the parameter estimation. Experimental data show that the HMM-based clustering procedure leads to $1.43\%$ performance improvements over the conventional template-based clustering method and $2.08\%$ improvements over the single HMM method for the case of recognition of the isolated korean digits.

• International journal of advanced smart convergence
• /
• v.7 no.2
• /
• pp.33-37
• /
• 2018

In this study, we propose an iterative clustering algorithm that automatically clusters a set of voice signal data without a label into an optimal number of clusters and generates hmm model for each cluster. In the clustering process, the likelihood calculations of the clusters are performed using iterative hmm learning and testing while varying the number of clusters for given data, and the maximum likelihood estimation method is used to determine the optimal number of clusters. We tested the effectiveness of this clustering algorithm on a small-vocabulary digit clustering task by mapping the unsupervised decoded output of the optimal cluster to the ground-truth transcription, we found out that they were highly correlated.

• The Journal of the Acoustical Society of Korea
• /
• v.32 no.2
• /
• pp.174-180
• /
• 2013

This paper proposes a decision tree based context clustering algorithm for HMM-based speech synthesis systems using the cross likelihood ratio with a hierarchical prior (CLRHP). Conventional algorithms tie the context-dependent HMM states that have similar statistical characteristics, but they do not consider the statistical similarity of split child nodes, which does not guarantee the statistical difference between the final leaf nodes. The proposed CLRHP algorithm improves the reliability of model parameters by taking a criterion of minimizing the statistical similarity of split child nodes. Experimental results verify the superiority of the proposed approach to conventional ones.

• The Journal of the Acoustical Society of Korea
• /
• v.18 no.3
• /
• pp.35-42
• /
• 1999

In this paper, a modified clustering algorithm is proposed to improve the discrimination of discrete HMM(Hidden Markov Model), so that it has increased recognition rate of 2.16% in comparison with the original HMM using the K-means or LBG algorithm. And, for preventing the decrease of recognition rate because of insufficient training data at the training scheme of HMM, a modified probabilistic smoothing method is proposed, which has increased recognition rate of 3.07% for the speaker-independent case. In the experiment applied the two proposed algorithms, the average rate of recognition has increased 4.66% for the speaker-independent case in comparison with that of original VQ/HMM.

• Journal of KIISE:Software and Applications
• /
• v.36 no.9
• /
• pp.683-690
• /
• 2009

One of the major problems that prevent us from building a good recognition system for large-scale on-line Chinese character recognition using HMMs is increasing recognition time. In this paper, we propose a clustering method to solve recognition speed problem and an efficient distance measure between HMMs. From the experiments, we got about twice the recognition speed and 95.37% 10-candidate recognition accuracy, which is only 0.9% decrease, for 20,902 Chinese characters defined in Unicode CJK unified ideographs.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1998.06c
• /
• pp.95-98
• /
• 1998

본 논문에서는 이산적인 HMM(Hidden Markov Model)을 이용한 고립단어 인식 시스템에서 입력특징 벡터의 변별력을 향상시키기 위해 수정된 집단화 알고리듬을 제안하므로써 K-means나 LBG 알고리듬을 이용한 기존의 HMM에 비해 2.16%의 인식율을 향상시켰다. 또한 HMM학습과정에서 불충분한 학습데이타로 인해 발생되는 인식율저하의 문제를 해소하기 위해 개선된 smoothing 기법을 제안하므로써 화자독립 실험에서 3.07%의 인식율을 향상시켰다. 본 논문에서 제안한 두가지 알고리듬을 모두 적용하여 최종적으로 실험한 VQ/HMM에서는 기존의 방식에 비해 화자독립 인식실험 결과 평균 인식율이 4.66% 개선되었다.

• Journal of Information Technology Application
• /
• v.1 no.3_4
• /
• pp.1-23
• /
• 1999

본 연구에서는 HMM(Hidden Markov Model) 및 Levelbuilding 알고리즘을 이용하여 인식대상 음소열의 표본 집합(훈련패턴 집합)을 입력으로 하는 음성의 자동 분할 시스템을 구현하였다. 본 시스템은 자연스럽게 발음되어진 연결음 음성으로부터 표준 음소모델을 생성한다. 본 시스템의 구성은 초기화 과정, HMM학습과정 그리고 Levelbuilding을 이용한 분리 및 CLustering 과정으로 구성되어 있다. 초기화 과정에서는 제어 정보를 이용하여 훈련패턴 집합으로부터 초기 음소 집합 군을 생성한다. Levelbuilding을 이용한 분리 및 Clustering 단계에서는 음소 모델과 제어 정보를 이용하여 훈련패턴들을 음소 단위로 분리하고, 분리된 후보 음소들을 Clustering하여 음소집합 군을 생성한다. 음소모델의 구성에 변화가 없을 때까지 이 작업을 반복 수행하여 최적의 음소모델을 생성한다. 본 연구에서는 3개 이하의 숫자단어로 구성된 연결되어 음성 패턴을 대상으로 실험하였다. 연결단어에 대한 음소의 표준모델 생성과정에서 가장 중요한 처리인 훈련패턴의 자동분할 과정을 분석하기 위하여 각 반복과정에서 분리된 정보를 그래프로 도시화하여 확인하였다.

• Proceedings of the KSPS conference
• /
• 2003.05a
• /
• pp.127-130
• /
• 2003

This paper describes an efficient method for unsupervised speaker adaptation. This method is based on selecting a subset of speakers who are acoustically close to a test speaker, and calculating adapted model parameters according to the previously stored sufficient HMM statistics of the selected speakers' data. In this method, only a few unsupervised test speaker's data are required for the adaptation. Also, by using the sufficient HMM statistics of the selected speakers' data, a quick adaptation can be done. Compared with a pre-clustering method, the proposed method can obtain a more optimal speaker cluster because the clustering result is determined according to test speaker's data on-line. Experiment results show that the proposed method attains better improvement than MLLR from the speaker independent model. Moreover the proposed method utilizes only one unsupervised sentence utterance, while MLLR usually utilizes more than ten supervised sentence utterances.

• Genomics & Informatics
• /
• v.21 no.3
• /
• pp.40.1-40.11
• /
• 2023

Microbial community profiling using 16S rRNA amplicon sequencing allows for taxonomic characterization of diverse microorganisms. While amplicon sequence variant (ASV) methods are increasingly favored for their fine-grained resolution of sequence variants, they often discard substantial portions of sequencing reads during quality control, particularly in datasets with large number samples. We present a streamlined pipeline that integrates FastP for read trimming, HmmUFOtu for operational taxonomic units (OTU) clustering, Vsearch for chimera checking, and Kraken2 for taxonomic assignment. To assess the pipeline's performance, we reprocessed two published stool datasets of normal Korean populations: one with 890 and the other with 1,462 independent samples. In the first dataset, HmmUFOtu retained 93.2% of over 104 million read pairs after quality trimming, discarding chimeric or unclassifiable reads, while DADA2, a commonly used ASV method, retained only 44.6% of the reads. Nonetheless, both methods yielded qualitatively similar β-diversity plots. For the second dataset, HmmUFOtu retained 89.2% of read pairs, while DADA2 retained a mere 18.4% of the reads. HmmUFOtu, being a closed-reference clustering method, facilitates merging separately processed datasets, with shared OTUs between the two datasets exhibiting a correlation coefficient of 0.92 in total abundance (log scale). While the first two dimensions of the β-diversity plot exhibited a cohesive mixture of the two datasets, the third dimension revealed the presence of a batch effect. Our comparative evaluation of ASV and OTU methods within this streamlined pipeline provides valuable insights into their performance when processing large-scale microbial 16S rRNA amplicon sequencing data. The strengths of HmmUFOtu and its potential for dataset merging are highlighted.

• Proceedings of the Korean Information Science Society Conference
• /
• 2008.06c
• /
• pp.472-477
• /
• 2008

본 논문에서는 은닉 마르코프 모델(HMM)을 기반한 대용량의 필기 한자 인식의 문제점인 시스템 리소스의 한계와 인식에 소요되는 많은 시간을 단축하기 위해 구조코드와 HMM에 최적화 된 클러스터링 알고리즘을 제안한다. 제안하는 클러스터링 알고리즘의 기본 개념은 훈련된 HMM를 대상으로 하고, HMM의 파라미터 수가 동일한 클래스에 대해서 클러스터를 구성하는 것이다. 또한 인식에 소요되는 시간을 줄이기 위해 2단계 클러스터모델 구조를 사용한다. 총 98,639 종류의 일본 한자를 대상으로 한 실험에서 평균 0.92 sec/char 인식 속도와 30순위 후보인식률 96.03%를 보임으로서 대용량 필기 한자 인식을 위한 좋은 방안이 될 것이라 기대한다.