• Proceedings of the Korean Society for Emotion and Sensibility Conference
• /
• 2009.05a
• /
• pp.181-182
• /
• 2009

GRBAS(G : grade, R : rough, B : breathy, S : strained, A : asthenic) 음성장애평가는 성대의 이상 또는 말마비장애 등의 환자들을 평가하는 척도로 널리 사용된다. 하지만 사람에 의해 주관적인 평가로 이루어지는 방식의 문제점이 많이 제기되어, 자동화 알고리즘에 의한 객관적인 청지각적 음성장애 평가도구를 개발하려는 시도가 많이 연구되어왔다. 이러한 개발에 있어 보편적으로 선행되어야 하는 음소 분류 및 일치성 판단을 위한 객관적인 파라미터를 구하고자 함이 본 연구의 목적이다.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2017.06a
• /
• pp.105-106
• /
• 2017

본 논문에서는 최대 우도 기반 파라미터 생성 알고리즘을 적용하여 인공 신경망의 출력인 음향 파라미터 열의 정확성 및 자연성을 향상시키는 방법을 제안하였다. 인공 신경망의 출력으로 정적 특징벡터 뿐 만 아니라 동적 특징벡터도 함께 사용하였고, 미리 계산된 파라미터 분산을 파라미터 생성에 사용하였다. 추정된 정적, 동적 특징벡터의 평균, 분산을 EM 알고리즘에 적용하여 최대 우도 기준 파라미터를 추정할 수 있다. 제안된 알고리즘은 파라미터 생성 시 동적 특징벡터 및 분산을 함께 적용하여 시간축에서의 자연성을 향상시켰다. 제안된 알고리즘의 객관적 평가로 MCD, F0 의 RMSE 를 측정하였고, 주관적평가로 선호도 평가를 실시하였다. 그 결과 기존 알고리즘 대비 객관적, 주관적 성능이 향상되는 것을 검증하였다.

• The Journal of the Acoustical Society of Korea
• /
• v.21 no.6
• /
• pp.531-542
• /
• 2002

Novel tandemless transcoding method for AMR and EVRC speech coders is proposed in this paper. In contrast to conventional tandem method, the parameters which is used commonly in speech coder where CELP algorithm is adapted are directly transcoded. The proposed algorithm is composed of LSP transcoding, pitch delay transcoding, gains transcoding and fixed codebook vector transcoding Evaluation results show that the novel algorithm achieves better speech quality than tandem method and reduce computational complexity and delay.

• Phonetics and Speech Sciences
• /
• v.16 no.1
• /
• pp.67-76
• /
• 2024

Recent advances in text-to-speech (TTS) technology have significantly improved the quality of synthesized speech, reaching a level where it can closely imitate natural human speech. Especially, TTS models offering various voice characteristics and personalized speech, are widely utilized in fields such as artificial intelligence (AI) tutors, advertising, and video dubbing. Accordingly, in this paper, we propose a one-shot multi-speaker TTS system that can ensure acoustic diversity and synthesize personalized voice by generating speech using unseen target speakers' utterances. The proposed model integrates a speaker encoder into a TTS model consisting of the FastSpeech2 acoustic model and the HiFi-GAN vocoder. The speaker encoder, based on the pre-trained RawNet3, extracts speaker-specific voice features. Furthermore, the proposed approach not only includes an English one-shot multi-speaker TTS but also introduces a Korean one-shot multi-speaker TTS. We evaluate naturalness and speaker similarity of the generated speech using objective and subjective metrics. In the subjective evaluation, the proposed Korean one-shot multi-speaker TTS obtained naturalness mean opinion score (NMOS) of 3.36 and similarity MOS (SMOS) of 3.16. The objective evaluation of the proposed English and Korean one-shot multi-speaker TTS showed a prediction MOS (P-MOS) of 2.54 and 3.74, respectively. These results indicate that the performance of our proposed model is improved over the baseline models in terms of both naturalness and speaker similarity.

• Proceedings of the Acoustical Society of Korea Conference
• /
• 1995.06a
• /
• pp.164-167
• /
• 1995

RPE 음성부호화기에서 합성 필터로 인한 구동벡터 양자화잡음의 증폭효과를 분석하고 regular pulse 시퀀스의 양자화로 인한 성능감쇄를 줄이기 위해 pyramid vector 양자화방식을 도입하였다. 제안된 방식의 성능평가는 구동시퀀스 양자화를 위해 adaptive PCM을 이용하는 GSM 표준 RPE 방식과의 객관적 및 주관적 성능비교를 통해 수행하였다.T JDSMDQLRY 결과 제안된 방식은 대략 1dB의 SNR 및 segmental SNR 값 증가를 가져왔고, 또한 비공식 청취시험결과 명료도의 증가를 느낄 수 있었다.

• Journal of rehabilitation welfare engineering & assistive technology
• /
• v.1 no.1
• /
• pp.21-28
• /
• 2007

In most of hearing aids, amplification algorithms are used to compensate hearing loss, noise and feedback reduction algorithms are used and to increase the perception of speeches contrast enhancement algorithms are used. However, acoustic masking effect is occurred between formants if contrast is enhanced excessively. To confirm the masking effect in speeches, the experiment are composed of 6 tests; test pure tone test, speech reception test, word recognition test, pure tone masking test, formant pure tone masking test and speech masking test, and for objective evaluation, LLR is introduced. As a result of normal hearing subjects and hearing impaired subjects, more making is occurred in hearing impaired subjects than normal hearing subjects when using pure tone, and in the speech masking test, speech reception is also lower in hearing impaired subjects than in normal hearing subjects. This means that acoustic masking effect rather than distortion influences speech perception. So it is required to check the characteristics of masking effect before wearing a hearing aid and to apply this characteristics to fitting curve.

• Science of Emotion and Sensibility
• /
• v.13 no.1
• /
• pp.111-120
• /
• 2010

Initial stage of emotion has a great influence on building up product image and impression. Because of its influencing effects, measuring initial stage of emotion has potential to be a key factor for designers and marketers to achieve a distinct product concept. While many researchers have studied this topic with the emotion measurement method in product use stage, there are very few cases specialized in the initial stage of emotion. Even though present emotion measurement methods have difficulties to derive accurate user's initial stage of emotion, most case of initial emotion study applies these defective methods. The purpose of this study is to develop initial stage of emotion measurement method and apply this method to real product context. In the design of the initial stage of emotion measurement method, noticeable characteristics of initial stage of emotion were explored and initial emotion measurement framework was presented. Based on this framework, Initial Emotion Measurement System(IEMS) was suggested. This method collects user's eye movement, behavior and verbal data accurately and objectively.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.49 no.3
• /
• pp.89-97
• /
• 2012

In this paper, we propose a novel approach to noise power estimation for speech enhancement in noisy environments. The method based on IMCRA (improved minima controlled recursive averaging) which is widely used in speech enhancement utilizes a rough VAD (voice activity detection) algorithm which excludes speech components during speech periods in order to improves the performance of the noise power estimation by reducing the speech distortion caused by the conventional algorithm based on the minimum power spectrum derived from the noisy speech. However, since the VAD algorithm is not sufficient to distinguish speech from noise at non-stationary noise and low SNRs (signal-to-noise ratios), the speech distortion resulted from the minimum tracking during speech periods still remained. In the proposed method, minimum power estimate obtained by IMCRA is modified by SMT (spectral minima tracking) to reduce the speech distortion derived from the bias of the estimated minimum power. In addition, in order to effectively estimate minimum power by considering the distribution characteristic of the speech and noise spectrum, the presented method combines the minimum estimates provided by IMCRA and SMT depending on the weighting factor based on the subband. Performance of the proposed algorithm is evaluated by subjective and objective quality tests under various environments and better results compared with the conventional method are obtained.

• The Journal of the Acoustical Society of Korea
• /
• v.16 no.1
• /
• pp.10-15
• /
• 1997

In this paper, an efficient LSF quantizer, named improved PSVQ(IPSVQ), is proposed to apply in the 8 kbps QCELP speech coder. By using 27 bits IPSVQ instead of 40 bits DPCM quantizer per frame, we can save 13 bits/frame and allocate those bits to the codebook gain and the pitch gain parameters. Hence we improve the overall performance of the QCELP codec. The enhanced QCELP shows the performance improvement of 0.9 dB SNR and 0.4 dB SEGSNR. Informal listening tests also confirm the improvement in the speech quality.

• Journal of the Korean Society of Laryngology, Phoniatrics and Logopedics
• /
• v.22 no.2
• /
• pp.137-142
• /
• 2011

Background and Objectives : The aim of this study is to introduce Korea version of pediatric VHI and to compare pVHI-K scores between children with dysphonia and children without voice problems before pVHI-K is developed as a preliminary study. Additionally, the relationship between pVHI and acoustic measures were investigated. Materials and Methods : pVHI-K scores in normal group were obtained from 15 parents who have children with no present or past history of a voice disorder, hearing loss, or related disability that can affect the their voice or speech. Dysphonia group consisted of 15 parents who have children with bilateral vocal fold nodule's at Department of Otolaryngology, the Seoul National University Hospital (SNUH). pVHI-K and acoustic parameters were measured in two group. Results : The mean pVHI scores (total, functional, physical, emotional) in normal group were 2.33 (T), 0.80 (F) 1.33 (P) and 0.27 (E), respectively whereas those of pVHI in children group with dysphonia were 23.13 (T), 11.07 (F), 5.73 (P) and 6.13 (E), respectively and significant differences were revealed in total pVHI score as well as in all of the sub-pVHI scores. Moreover, significant correlation between pVHI-K parameters (T, F, P) and acoustic measures [Shimmer(%)] were shown in children in dysphonia group. Conclusion : Reported by parents can be useful as a supplementary clinical tool for diagnosing and measuring treatment effectiveness in young children with dysphonia.