• 대한음성학회지:말소리
• /
• 제52호
• /
• pp.111-120
• /
• 2004

This paper focuses on the DSP implementation of an HMM-based speech recognizer that can handle several hundred words of vocabulary size as well as speaker independency. First, we develop an HMM-based speech recognition system on the PC that operates on the frame basis with parallel processing of feature extraction and Viterbi decoding to make the processing delay as small as possible. Many techniques such as linear discriminant analysis, state-based Gaussian selection, and phonetic tied mixture model are employed for reduction of computational burden and memory size. The system is then properly optimized and compiled on the TMS320C6711 DSP for real-time operation. The implemented system uses 486kbytes of memory for data and acoustic models, and 24.5 kbytes for program code. Maximum required time of 29.2 ms for processing a frame of 32 ms of speech validates real-time operation of the implemented system.

• The Journal of the Acoustical Society of Korea
• /
• 제16권1E호
• /
• pp.57-63
• /
• 1997

In this paper, we investigated performance of a vocabulary-independent speech recognizer with speaker adaptation. The vocabulary-independent speech recognizer does not require task-oriented speech databases to estimate HMM parameters, but adapts the parameters recursively by using input speech and recognition results. The recognizer has the advantage that it relieves efforts to record the speech databases and can be easily adapted to a new task and a new speaker with different recognition vocabulary without losing recognition accuracies. Experimental results showed that the vocabulary-independent speech recognizer with supervised offline speaker adaptation reduced 40% of recognition errors when 80 words from the same vocabulary as test data were used as adaptation data. The recognizer with unsupervised online speaker adaptation reduced abut 43% of recognition errors. This performance is comparable to that of a speaker-independent speech recognizer trained by a task-oriented speech database.

• 대한의용생체공학회:학술대회논문집
• /
• 대한의용생체공학회 1992년도 춘계학술대회
• /
• pp.218-218
• /
• 1992

• 한국통신학회논문지
• /
• 제23권6호
• /
• pp.1443-1453
• /
• 1998

본 논문은 컴퓨터에 입력된 문자정보를 음성정보로 변환하기 위한 음성합성엔진에 관한 것이며, 특히 명료성의 향상을 위해 파형처리 음성합성방식을 이용한다. 음성합성엔진은 컴맨드 스트림의 제어에 따라 자연성의 향상을 위한 피치조절, 길이 및 에너지 등을 제어하며 음성합성단위로서 반음절을 사용한다. 엔진에서 사용 가능한 컴맨드를 프로그램하여 음성합성엔진에 입력함으로서 음성을 합성하는 빙식은 구문분석, 어휘분석 등의 하이레벨과 파형의 편집 가공 등의 로우레벨을 완전 분리하므로 시스템의 융통성과 확장성을 높인다. 또한 TTS시스템의 적용에 있어 각 모듈을 객체/컴포넌트(Object/Component)로 각 모듈이 상호 독립적으로 작동되도록 하여 쉽게 대체가 가능하다. 하이 레벨과 로우 레벨을 분리하는 소프트웨어 아키택처는 음성합성 연구에 있어 각각 여러 분야별로 독립적으로 연구수행이 가능하여 연구의 효율성을 높이며 여러 소프트웨어의 조합사용(Mix-and-Match)이 가능하여 확장성과 이식성을 향상시킨다.

• 대한음성학회지:말소리
• /
• 제57호
• /
• pp.99-112
• /
• 2006

In this paper, we introduce two retrieval methods for photos with speech documents. We compare the pattern of speech query with those of speech documents recorded in digital cameras, and measure the similarities, and retrieve photos corresponding to the speech documents which have high similarity scores. As the first approach, a phoneme recognition scheme is used as the pre-processor for the pattern matching, and in the second one, the vector quantization (VQ) and the dynamic time warping (DTW) are applied to match the speech query with the documents in signal domain itself. Experimental results show that the performance of the first approach is highly dependent on that of phoneme recognition while the processing time is short. The second method provides a great improvement of performance. While the processing time is longer than that of the first method due to DTW, but we can reduce it by taking approximated methods.

• 말소리와 음성과학
• /
• 제2권1호
• /
• pp.77-85
• /
• 2010

This paper proposes an efficient feature vector processing technique to guard the Speech Emotion Recognition (SER) system against a variety of noises. In the proposed approach, emotional feature vectors are extracted from speech processed by comb filtering. Then, these extracts are used in a robust model construction based on feature vector classification. We modify conventional comb filtering by using speech presence probability to minimize drawbacks due to incorrect pitch estimation under background noise conditions. The modified comb filtering can correctly enhance the harmonics, which is an important factor used in SER. Feature vector classification technique categorizes feature vectors into either discriminative vectors or non-discriminative vectors based on a log-likelihood criterion. This method can successfully select the discriminative vectors while preserving correct emotional characteristics. Thus, robust emotion models can be constructed by only using such discriminative vectors. On SER experiment using an emotional speech corpus contaminated by various noises, our approach exhibited superior performance to the baseline system.

• 한국정보처리학회논문지
• /
• 제4권6호
• /
• pp.1531-1540
• /
• 1997

본 논문에서는 발화속도 측정과 이를 통한 보상방법을 통하여 성능 향상된 한국어 연속음 인식 시스템을 제안한다. 연속음 인식은 다양한 조음화 현상과 발화속도의 변화로 인하여 고립단어 인식에 비하여 어렵다. 따라서, 연속음 인식을 위해서는 조음화 현상과 발화속도의 변화를 모델링할 수 있는 방법이 필요하다. 본 논문에서는 발화속도를 포만트의 변화율로서 측정하였고, 이 정보를 이용하여 빠른 발화에서는 상대적으로 많은 특징벡터를 발생시켜 보상을 시도하였다. 또한 조음화 현상을 모델링하기 위하여 한국어의 다이폰 집합을 514개로 정의하였고, 훈련을 위한 음성 DB론느 ETRI의 445 단어 DB를 사용하였다. 이러한 방법을 결합한 한국어 연속음 인식기를 DHMM (Discrete Hidden Markov Model)으로 구현하여 인식률이 향상됨을 보였다.

• 한국산업정보학회논문지
• /
• 제7권3호
• /
• pp.74-80
• /
• 2002

한국어 연속 음성 인식결과의 성능향상을 위해서 자연어 처리 기술을 이용한 후처리 기법이 사용된다. 그러나 자연어 처리 기법이 대부분 띄어쓰기가 있는 정형화된 입력 문장에 대한 분석을 수행하여 왔기 때문에 형태소 분석기를 직접 음성인식 결과의 향상에 사용하는 데에는 어려운 점이 많다. 본 논문에서는 띄어쓰기를 고려하지 않는 기능어 기반의 최장일치 형태소 해석 방법인 구문 형태소 단위의 분석을 이용한 음정인식 결과의 향상 모델을 제안한다. 제안된 모델을 통해 연속음성 인식 결과에서 자주 발생하는 용언과 보조 용언 및 의존 명사 사이의 음운들 사이의 구조적 정보를 활용함으로써 음성 인식 결과의 성능을 향상시키는 방법에 대해 기술한다.

• 말소리와 음성과학
• /
• 제9권3호
• /
• pp.57-65
• /
• 2017

The current study explored how Korean-speaking children processed the multiple acoustic cues (VOT and f0) for the stop laryngeal contrast (/t'/, /t/, and /$t^h$/) and examined whether individual perceptual strategies could be related to a general cognitive ability performing executive functions (EF). 15 children (aged from 7 to 8) participated in the speech perception task identifying the three Korean laryngeal stops (3AFC) on listening to the auditory stimuli of C-/a/ with synthetically varying VOT and f0. They completed a series of EF tasks to measure working memory, inhibition, and cognitive shifting ability. The findings showed that children used the two cues in a highly correlated manner. While children utilized VOT consistently for the three laryngeal categories, their use of f0 was either reduced or enhanced depending on the phonetic categories. Importantly, the children's processing strategies of a f0 suppression for a tense-aspirated contrast were meaningfully associated with children's better cognitive abilities such as working memory, inhibition, and attentional shifting. As a preliminary experimental investigation, the current research demonstrated that listeners with inefficient processing strategies were poor at the EF skills, suggesting that cognitive skills might be responsible for developmental variations of processing sub-phonemic information for the linguistic contrast.

• 대한음성학회지:말소리
• /
• 제50호
• /
• pp.41-50
• /
• 2004

Recently, there has been considerable progress in the natural language processing and digital signal processing components and this progress has led to the improved synthetic speech qualify of many commercial TTS systems. But there still remain many obstacles to overcome for the practical application of TTS. To resolve the problems, the cooperative research among the related areas is highly required and a common Korean TTS platform is essential to promote these activities. This platform offers a general framework for building Korean speech synthesis systems and a full C/C++ source for modules supports to implement and test his own algorithm. In this paper we described the aspect of a Korean TTS platform to be developed and a developing plan.