• Annual Conference on Human and Language Technology
• /
• 2014.10a
• /
• pp.9-12
• /
• 2014

본 논문에서는 오인식된 고유명사의 후처리 방법을 제안한다. 최근 음성 인식 후처리를 위해 통계적 방법을 이용하는 연구가 활발히 진행되어 왔다. 하지만 고유명사의 음성 인식 후처리는 대용량의 데이터 수집에 많은 비용이 필요하므로 통계적 방법을 효과적으로 적용하기 어렵다. 따라서 본 논문에서는 발음 변이 현상을 고려하여 편집 거리 알고리즘을 개선한 기법을 제안한다. 본 논문에서는 고유명사의 음성 오인식 교정 성능을 검증하였고, 그 결과 P@3의 결과가 비교 모델보다 55%의 성능 향상률을 보였다.

• The Journal of the Acoustical Society of Korea
• /
• v.14 no.5
• /
• pp.44-50
• /
• 1995

This paper describes a large-vocabulary continuous speech recognition system using continuous hidden Markov models for the Korean language. To improve the performance of the system, we study on the selection of speech modeling units, inter-word modeling, search algorithm, and grammars. We used triphones as basic speech modeling units, generalized triphones and function word-dependent phones are used to improve the trainability of speech units and to reduce errors in function words. Silence between words is optionally inserted by using a silence model and a null transition. Word pair grammar and bigram model based oil word classes are used. Also we implement a search algorithm to find N-best candidate sentences. A postprocessor reorders the N-best sentences using word triple grammar, selects the most likely sentence as the final recognition result, and finally corrects trivial errors related with postpositions. In recognition tests using a 3,000-word continuous speech database, the system attained $93.1\%$ word recognition accuracy and $73.8\%$ sentence recognition accuracy using word triple grammar in postprocessing.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1998.08a
• /
• pp.257-261
• /
• 1998

일반적으로 인식률이 저조한 연속 숫자음의 인식 정도 향상을 위해서 K-L 동적특징의 확장에 대해서 검토한다. 이 검토결과를 4연속 숫자음을 대상으로 하는 인식 실험을 수행하여 숫자음 인식에 있어서 확장된 K-L 동적특징의 유효성을 확인하고자 한다. 이를 위하여 음성자료는 국어공학센터에서 채록한 4연속 숫자음을 사용하며, 확장한 K-L 동적특징의 유효성을 확인하기 위해서는 단일 특징 파라미터로서 멜-켑스트럼과 회귀계수, K-L 동적계수 등과 이들 특징 파라미터를 결합한 경우에 대해서 특징파라미터를 확장하여 K-L 동적 특징을 추출하고, 4연속 숫자음인식 실험을 수행하였다. 이때 인식의 기본 단위로는 48개의 유사음소단위를 음소모델로 사용하였으며, 인식실험에 있어서는 유한 상태 오토마타에 의한 구문제어를 통한 OPDP 법을 이용하였다. 인식 실험 결과, 단일 특징파라미터로서 멜-켑스트럼을 사용한 경우 67.5%, 이를 확장한 K-L 동적계수를 사용한 경우 78.2%를 보였다. 또한 결합한 특징파라미터에 있어서는 멜-켑스트럼과 희귀계수를 사용한 경우 78.4%의 인식률을 보였으며, 이를 K-L 동적계수로 확장한 경우 82.3%의 인식률을 얻어 확장한 K-L 동적특징파라미터의 유효성을 확인하였다.

• Proceedings of the Korea Multimedia Society Conference
• /
• 2003.05b
• /
• pp.524-527
• /
• 2003

최근 들어 립리딩은 멀티모달 인터페이스 기술의 응용분야에서 많은 관심을 모으고 있다. 동적 영상을 이용한 립리딩 시스템에서 해결해야 할 주된 문제점은 상황 변화에 독립적인 얼굴 영역과 입술 영역을 추출하는 것이다. 본 논문에서는 움직임이 있는 영상에서 화자의 얼굴영역과 입술영역을 컬러, 조명등의 변화에 독립적으로 추출하기 위해 HSI 모델과 블록 매칭을 이용하였고 특징 점 추출에는 이미지 기반 방법인 PCA 기법을 이용하였다. 추출된 입술 파라미터와 음성 데이터에 각각 HMM 기반 패턴 인식 방법을 개별적으로 적용하여 단어를 인식하였고 각각의 인식 결과를 가중치를 주어 합병하였다. 실험 결과에 의하면 잡음으로 음성 인식률이 낮아지는 경우에 음성인식과 립리딩을 함께 사용함으로써 전체적인 인식 결과를 향상시킬 수 있었다.

• The Journal of the Acoustical Society of Korea
• /
• v.12 no.2E
• /
• pp.5-11
• /
• 1993

본 논문에서 HMM 파라메터 추정을 위해 분절단위 정보를 이용하는 수정된 교정학습방법을 제안한다. 수정된 교정학습방법은 기존의 교정학습 방법에서 사용하는 전향·후향 알고리즘 대신에 분절단위 K-means 알고리즘을 사용하여 HMM 파라메터를 교정한다. 이 방식은 분절단위 K-means 알고리즘이 음성신호내의 공통의 통계적 특성을 가지는 상태단위 정보를 강조한다는 사실을 이용하였다. 화자종속 음소 및 단어인식 실험에서 제안된 알고리즘이 기존의 교정학습 방법보다 적은 계산량으로도 향상된 인식률을 보여주었다. 이것은 HMM 교정학습에서 상태다누이 정보가 중요함을 보여준다.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.17 no.2
• /
• pp.280-286
• /
• 2013

In this paper, a noise reduction system is proposed for the speech interface in intelligent TV applications. To reduce TV speaker sound which are very serious noises degrading recognition performance, a noise reduction algorithm utilizing the direct TV sound as the reference noise input is implemented. In the proposed algorithm, transfer functions are estimated to compensate for the difference between the direct TV sound and that recorded with the microphone installed on the TV frame. Then, the noise power spectrum in the received signal is calculated to perform Wiener filter-based noise cancellation. Additionally, a postprocessing step is applied to reduce remaining noises. Experimental results show that the proposed algorithm shows 88% recognition rate for isolated Korean words at 5 dB input SNR.

• Smart Media Journal
• /
• v.11 no.10
• /
• pp.22-29
• /
• 2022

Voice interactions are particularly effective in applications targeting the digital underprivileged who are not proficient in the use of smart devices. However, applications based on open APIs are using voice signals only for short, fragmentary input and output due to the limitations of existing touchscreen-oriented UI and API provided. In this paper, we design a conversational voice interaction model for interactions between users and intelligent mobile/IoT applications and propose a keyword detection algorithm based on the edit distance. The proposed model and scheme were implemented in an Android environment, and the edit distance-based keyword detection algorithm showed a higher recognition rate than the existing algorithm for keywords that were incorrectly recognized through speech recognition.

• Journal of Korea Multimedia Society
• /
• v.13 no.4
• /
• pp.496-503
• /
• 2010

Speech recognition system is developed of standalone, In case of a mobile terminal using that low recognition rate represent because of limitation of memory size and audio compression. This study suggest vocabulary recognition highest performance improvement system for separate acoustic search and lexical search. Acoustic search is carry out in mobile terminal, lexical search is carry out in server processing system. feature vector of speech signal extract using GMM a phoneme execution, recognition a phoneme list transmission server using Lexical Tree Search algorithm lexical search recognition execution. System performance as a result of represent vocabulary dependence recognition rate of 98.01%, vocabulary independence recognition rate of 97.71%, represent recognition speed of 1.58 second.

• The Journal of the Acoustical Society of Korea
• /
• v.26 no.6
• /
• pp.250-258
• /
• 2007

Current speech recognition technology s achieved high performance with the development of hardware devices, however it is insufficient for some applications where high reliability is required, such as voice control of powered wheelchairs for disabled persons. For the system which aims to operate powered wheelchairs safely by voice in real environment, we need to consider that non-voice commands such as user s coughing, breathing, and spark-like mechanical noise should be rejected and the wheelchair system need to recognize the speech commands affected by disability, which contains specific pronunciation speed and frequency. In this paper, we propose non-voice rejection method to perform voice/non-voice classification using both YIN based fundamental frequency(F0) extraction and reliability in preprocessing. We adopted a multi-template dictionary and acoustic modeling based speaker adaptation to cope with the pronunciation variation of inarticulately uttered speech. From the recognition tests conducted with the data collected in real environment, proposed YIN based fundamental extraction showed recall-precision rate of 95.1% better than that of 62% by cepstrum based method. Recognition test by a new system applied with multi-template dictionary and MAP adaptation also showed much higher accuracy of 99.5% than that of 78.6% by baseline system.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.35S no.12
• /
• pp.93-102
• /
• 1998

In noisy environments, human speech productions are influenced by noises(Lombard effect), and speech signals are contaminated. These distortions dramatically reduce the performance of speech recognition systems. This paper proposes a method of the Lombard effect compensation and noise suppression in order to improve speech recognition performance in noise environments. To estimate the intensity of the Lombard effect which is a nonlinear distortion depending on the ambient noise levels, speakers, and phonetic units, we formulate the measure of the Lombard effect level based on the acoustic speech signal, and the measure is used to compensate the Lombard effect. The distortions of speech under noisy environments are cancelled out as follows. First, spectral subtraction and band-pass filtering are used to cancel out noise. Second, energy nomalization is proposed to cancel out the variation of vocal intensity by the Lombard effect. Finally, the Lombard effect level controls the transform which converts Lombard speech cepstrum to clean speech cepstrum. The proposed method was validated on 50 korean word recognition. Average recognition rates were 82.6%, 95.7%, 97.6% with the proposed method, while 46.3%, 75.5%, 87.4% without any compensation at SNR 0, 10, 20 dB, respectively.