• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.24 no.6A
• /
• pp.820-831
• /
• 1999

In this paper, we have compared the performance of a D-TDMA protocol with that of a CDMA protocol, in radio access system for integrated voice/data services.The D-TDMA protocol is based on a generic dynamic channel assignment approach to be followed a combination of “circuit mode” reservation for voice calls, along with dynamic first-come-first served assignment of remaining capacity for data messages. In the CDMA protocol, we have used the voice activity detection to reduce the interface power of other mobiles in internal and external cells, and analyzed the interference power ratio. Also we have computed BER(Bit Error Rate) by using this interference power ratio and evaluated voice blocking probability(voice packet loss probability) and data transmission delay, according to average data length and average data arrival rate.We have found the CDMA protocol achieves comparatively higher performance for short data length, regardless of data arrival rate. Otherwise, the data transmission delay of D-TDMA protocol is shorter than that of the CDMA protocol for long data message.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.30 no.9A
• /
• pp.737-748
• /
• 2005

Two flexible admission control schemes for integrated voice and stream-type data services are proposed in DS-CDMA systems. Most Previous studies on admission control have focused on integration of short, bursty Packet-type data services and conventional voice services. However, stream-type data services with a relatively long service holding time are expected to be a considerable portion of data traffic in future generation cellular systems. Scheme I is a basic scheme that accommodates both voice and data services with full bandwidth. However, voice services are given priority over data services using the duration difference between the holding times for these services. Scheme ll uses a different method to efficiently give priority to voice services over stream-type data services. An additional interference margin for voice services is provided by suppressing interference from stream-type data services according to voice access requests and a varying interference status. Performance of the two schemes is evaluated by developing Markovian models. Numerical results show that the voice capacity is highly sensitive to the service holding time of data services while the performance measures of data services are not highly sensitive. Scheme H is a significant improvement over Scheme I for accommodating voice and stream-type data services

• The Korean Journal of Applied Statistics
• /
• v.34 no.3
• /
• pp.449-460
• /
• 2021

ASVspoof 2017 deals with detection of replay attacks and aims to classify real human voices and fake voices. The spoofed voice refers to the voice that reproduces the original voice by different types of microphones and speakers. data augmentation research on image data has been actively conducted, and several studies have been conducted to attempt data augmentation on voice. However, there are not many attempts to augment data for voice replay attacks, so this paper explores how audio modification through data augmentation techniques affects the detection of replay attacks. A total of 7 data augmentation techniques were applied, and among them, dynamic value change (DVC) and pitch techniques helped improve performance. DVC and pitch showed an improvement of about 8% of the base model EER, and DVC in particular showed noticeable improvement in accuracy in some environments among 57 replay configurations. The greatest increase was achieved in RC53, and DVC led to an approximately 45% improvement in base model accuracy. The high-end recording and playback devices that were previously difficult to detect were well identified. Based on this study, we found that the DVC and pitch data augmentation techniques are helpful in improving performance in the voice spoofing detection problem.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.05a
• /
• pp.491-493
• /
• 2022

최근 스마트 기기에서 오디오 데이터를 이용하는 응용 기술들이 증가하면서, 오디오 데이터에서 관심 있는 구간을 찾아내는 기술의 필요성이 증가하고 있다. 본 논문에서는 Perceiver 모델을 활용하여 오디오 데이터에서 사람의 음성 구간을 검출하고 축약하는 방법을 제안한다. Perceiver 모델은 복잡한 입력 데이터에 대하여 Self-attention을 기반으로 특징을 추출하면서 이전의 특징을 다음 입력으로 다시 학습하는 특징을 갖고 있어서 연속적인 데이터인 오디오에 효율적으로 적용할 수 있다. 외부 및 자체에서 수집한 음성과 비음성 데이터셋에 대하여 실험을 진행하였고, 10초 단위 세그먼트에서 대해서 92.4%의 검출 정확도를 달성하였다.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.11 no.3
• /
• pp.179-196
• /
• 1986

In this paper, the peformance of a statistical packet voice/data multiplexer is studied. In ths study we assume that in the packet voice/data multiplexer two separate finite queues are used for voice and data traffics, and that voice traffic gets priority over data. For the performance analysis we divide the output link of the multiplexer into a sequence of time slots. The voice signal is modeled as an (M+1) - state Markov process, M being the packet generation period in slots. As for the data traffic, it is modeled by a simple Poisson process. In our discrete time domain analysis, the queueing behavior of voice traffic is little affected by the data traffic since voice signal has priority over data. Therefore, we first analyze the queueing behavior of voice traffic, and then using the result, we study the queueing behavior of data traffic. For the packet voice multiplexer, both inpur state and voice buffer occupancy are formulated by a two-dimensional Markov chain. For the integrated voice/data multiplexer we use a three-dimensional Markov chain that represents the input voice state and the buffer occupancies of voice and data. With these models, the numerical results for the performance have been obtained by the Gauss-Seidel iteration method. The analytical results have been verified by computer simylation. From the results we have found that there exist tradeoffs among the number of voice users, output link capacity, voic queue size and overflow probability for the voice traffic, and also exist tradeoffs among traffic load, data queue size and oveflow probability for the data traffic. Also, there exists a tradeoff between the performance of voice and data traffics for given inpur traffics and link capacity. In addition, it has been found that the average queueing delay of data traffic is longer than the maximum buffer size, when the gain of time assignment speech interpolation(TASI) is more than two and the number of voice users is small.

• 한국IT서비스학회:학술대회논문집
• /
• 2009.05a
• /
• pp.365-368
• /
• 2009

고속 인터넷 망 구축과 멀티미디어 통신 수요의 증가에 따라 VoIP는 기존의 PSTN 망의 대체 혹은 확장 기술로서 지속적으로 검증되어 왔다. 음성 데이터 처리 규약들 중 SIP는 다른 규약에 비해 신호 처리 단계가 간단하기 때문에 이를 기반으로 RTP를 활용하여 음성 통신 시스템을 구축하는 사례가 늘어나고 있다. 그러나 RTP의 특성상 패킷을 처리할 때마다 복원 과정이 필요하며, 다중 세션으로 통신이 발생할 경우 전체 패킷들의 관리가 복잡해지므로 이들 간에 혼선 없이 데이터를 처리 및 유지할 수 있는 방법론이 요구된다. 본 논문에서는 SIP 기반의 IP 전화를 통해서 고객과 상담원 간의 통화 이벤트가 발생하는 일반 콜센터 환경에서 RTP 음성 데이터를 처리하는 다중 세션 어플리케이션의 구축 사례를 제시한다. 구현한 시스템은 IP 전화에서 발생하는 통화 내역을 통합 스위치 서버에서 포트 미러링하여 녹취 및 녹음 서버로 전송하며, 전송된 패킷 정보들의 세션이 유지되고 있는 동안 음성 데이터를 실시간으로 모니터링한다.

• Proceedings of the Acoustical Society of Korea Conference
• /
• autumn
• /
• pp.175-178
• /
• 2000

GMM을 기반으로 하는 화자식별 시스템은 입력음성의 길이의 장단에 의해서 인식률에 차이가 생긴다. 이는 가우시안 모델의 파라미터를 추정할 때, 않은 데이터를 사용할수록 추정이 정확해지기 때문이다. 따라서 화자식별에 사용하는 입력데이터는 화자가 발성한 모든 음성신호에서 잡음구간만을 제거한 유,무성음을 이용하게 된다. 그러나 이 경우 데이터의 양이 많아져서 실시간 처리에 어려움이 있겠다. 본 논문에서는 전체 음성구간을 이용하는 대신 유성음 구간만을 추출하여 이 구간의 켑스트럼과 피치 값들을 특징파라미터로 이용하여 화자식별에 이용하였다. 특히 피치성분은 일반적으로 통신채널과 핸드셋의 영향에 상대적으로 강한 장점이 있다. 실험을 위하여 20대의 남성 및 여성화자 40명으로부터 얻은 음성데이터에서 유성음구간을 추출하여 GMM을 이용한 문장독립 화자식별 실험을 하였으며, 실험결과 스펙트럼정보와 함께 피치 정보가 화자식별에 유용하게 사용될 수 있음을 알 수 있었다

• The Journal of the Acoustical Society of Korea
• /
• v.41 no.6
• /
• pp.655-661
• /
• 2022

Since speech impairment is prevalent in patients with Parkinson's disease (PD), speech recognition systems suitable for these patients are needed. In this paper, we propose a speech recognition system that effectively recognizes the speech of patients with PD. The speech recognition system is firstly pre-trained with the Globalformer using the speech data from healthy people, and then fine-tuned using relatively small amount of speech data from the patient with PD. For this analysis, we used the speech dataset of healthy people built by AI hub and that of patients with PD collected at Inha University Hospital. As a result of the experiment, the proposed speech recognition system recognized the speech of patients with PD with Character Error Rate (CER) of 22.15 %, which was a better result compared to other methods.

• The Journal of the Acoustical Society of Korea
• /
• v.41 no.2
• /
• pp.122-129
• /
• 2022

In this paper, we propose a method to improve the performance of Transformer-based speech recognizers using data augmentation that locally adjusts the frame rate. First, the start time and length of the part to be augmented in the original voice data are randomly selected. Then, the frame rate of the selected part is changed to a new frame rate by using linear interpolation. Experimental results using the Wall Street Journal and LibriSpeech speech databases showed that the convergence time took longer than the baseline, but the recognition accuracy was improved in most cases. In order to further improve the performance, various parameters such as the length and the speed of the selected parts were optimized. The proposed method was shown to achieve relative performance improvement of 11.8 % and 14.9 % compared with the baseline in the Wall Street Journal and LibriSpeech speech databases, respectively.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 1999.11b
• /
• pp.141-145
• /
• 1999

뉴스나 드라마, 영화 등의 비디오에 대한 검색 시 일반 사용자의 요구에 가장 잘 부합되는 결과를 얻기 위해 비디오 데이터의 의미적 분석과 색인을 만드는 것이 필요하다. 일반적으로 음성신호가 비디오 데이터의 내용을 잘 나타내고 비디오와 동기가 이루어져 있으므로, 내용기반 검색을 위한 비디오 데이터 분할에 효율적으로 이용될 수 있다 본 논문에서는 캡션 정보가 주어지는 방송뉴스 프로그램을 대상으로 효율적인 검색, 색인을 위한 비디오 데이터의 분할에 음성인식기술을 적용하는 방법을 제안하고 그에 따른 실험결과를 제시한다.