• Proceedings of the KIEE Conference
• /
• 2002.11c
• /
• pp.605-608
• /
• 2002

A text-dependent speaker recognition system using a robust matching process is studied. The feature histogram of LPC cepstral coefficients for matching is used. The matching process uses mixture network with penalty scores. Using probability and shape comparison of two feature histograms, similarity values are obtained. The experiment results will be shown to show the effectiveness of the proposed algorithm.

• The Journal of the Acoustical Society of Korea
• /
• v.18 no.3E
• /
• pp.30-36
• /
• 1999

In this paper, we develop a hybrid speaker recognition system [1] enhanced by pre-recognizer and post-recognizer. The pre-recognizer consists of general speech recognition systems and the post-recognizer is a pitch detection system using adapted continuous wavelet transform (ACWT) to improve the performance of the hybrid speaker recognition system. Two schemes to design ACWT is considered. One is the scheme to search basis library covering the whole band of speech fundamental frequency (speech pitch). The other is the scheme to determine which one is the best basis. Information cost functional is used for the criterion for the latter. ACWT is robust enough to classify the pitch of speech very well, even though the speech signal is badly damaged by environmental noises.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.13 no.3
• /
• pp.17-24
• /
• 1999

This paper presents a method of ANC for transfonrer noise control utilizing a sproker and microphone pair. In this study, the main focus is on identifying the dynamic characteristics of speaker - amplifier microphone path. This study presents a theoretical method to identify the dynamic characteristics of speaker-microphone pairs. The transfer functions of microphone - speaker pair have been estimated utilizing sequential least square(SLS) algorithm. We identified the estimated transfer function has stable JXlles and zeros in z-plane. This paper also propose an architecture far the noise cancellation to which we applied the estimated transfer function.nction.

• Journal of Information Technology Services
• /
• v.17 no.2
• /
• pp.191-205
• /
• 2018

With the advancement of artificial intelligence (AI) techniques, many consumer products have adopted AI features for providing proactive and personalized services to customers. One of the most prominent products featuring AI techniques is a smart speaker. The fundamental of smart speaker is a portable wireless Internet connecting speaker which already have existed in a consumer market. By applying AI techniques, smart speakers can recognize human voices and communicate with them. In addition, they can control other connecting devices and provide offline services. The goal of this study is to identify the impact of AI techniques for customer rating to the products. We compared customer reviews of other portable speakers without AI features and those of a smart speaker. Amazon echo is used for a smart speaker and JBL Flip 4 Bluetooth Speaker and Ultimate Ears BOOM 2 Panther Limited Edition are used for the comparison. These products are in the same price range ($50~100) and selected as featured products in Amazon.com. All reviews for the products were collected and common words for all products and unique words of the smart speaker were identified. Information gain values were calculated to identify the influences of words to be rated as positive or negative. Positive and negative words in all the products or in Amazon echo were identified, too. Topic modeling was applied to the customer reviews on Amazon echo and the importance of each topic were measured by summating information gain values of each topic. This study provides a way of identifying customer responses on the AI feature and measuring the importance of the feature among diverse features of the products.

• Journal of English Language & Literature
• /
• v.56 no.4
• /
• pp.753-772
• /
• 2010

The purpose of this work is to examine the viability of employing the prosody cloning technique in teaching English prosody. Ten native speakers of Korean high school students with similar level of English proficiency participated in the poem self-study experiment. Five of them were grouped into the experimental group and the remaining five into the control group. One popular English poem from a high school textbook was selected and its recording by a professional native speaker of English was used in the experiment. The members of the two groups made a recording of the poem both before and after the experiment. For the study material, the experimental group used their own recorded utterances with their prosody cloned from the professional English speaker, while the control group used the utterances of the professional speaker alone. The acoustic analysis of the recordings by the prosodic foot both before and after the experiment showed that the experimental group performed slightly better than the control group in the realization of the intensity contour of the poem. There were no significant differences in the realization of the intonation contour and segmental durations between the two groups. The recording after the experiment was also subjectively evaluated by a native speaker of English and the scores for the experimental group were slightly higher than the control group. These findings suggest that the use of English poems with the help of the prosody cloning technique is a potentially viable approach to teaching English intonation to high school students. A long-term study with more students is necessary.

• Proceedings of the IEEK Conference
• /
• 1999.06a
• /
• pp.1095-1098
• /
• 1999

We have developed a speaker dependent voice command system(VCS) to control the sunroof in the car using RSC-164 VRP(Voice Recognition Processor). VCS consists of control circuits, microphone, speaker and user switch box. The control circuits include RSC-164, input audio preamplifier, memory devices, and relay circuit for sunroof control. It is designed robustly in various car noisy situations like audio volume, air conditioner, and incoming noise when window or sunroof opened. Each two users can control the car sunroof using seven voice commands on the Super TVS model and five voice commands on the Onyx model. It works well when we drive the car at over 100 km/h with the sunroof opened.

• Proceedings of the KIEE Conference
• /
• 2003.07d
• /
• pp.2766-2768
• /
• 2003

In this paper, positions of microphones and speakers in adaptive noise control systems are selected using optimization techniques. To get the optimal control characteristics of adaptive noise control system, It is necessary to optimize positions of microphones and speakers. Assume that control speakers are placed in a experimental enclosure several place indoors. Acoustic transfer functions from control speaker positions to microphone positions are measured for simulation program by experiments, and simulated annealing methods are used to select optimal positions of speakers and microphones.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2009.04a
• /
• pp.247-249
• /
• 2009

This paper investigates active noise control(ANC) of a rectangular cavity using single channel filtered-x least mean square(FXLMS) algorithms to reduce interior noise globally. To obtain global reduction of the interior noise, multichannel active control should be incorporated in general. We, however, examined firstly the optimal location of the secondary speaker that produces a global reduction of the interior noise field. We then investigated the frequency characteristics of the reduction to yield the effective frequency band of the active control system. It follows that the secondary speaker should be located as close to the primary source as possible in order to obtain global reduction.

• Phonetics and Speech Sciences
• /
• v.1 no.4
• /
• pp.171-176
• /
• 2009

Intelligibility measurement is one criterion for the assessment of the severity of speech disorders especially of dysarthric persons. Rate control, usually rate reduction, is used with many dysarthric speakers to improve their intelligibility. The purpose of this study is to compare how change intelligibility of speech produced by cerebral palsic speakers according to three speaking conditions. Speech samples were collected from 10 adults with cerebral palsy were asked to speak under three speaking conditions-(1) naturally(control), (2) more slowly(rate control), (3) louder and accurately(clear speech). In a perception test, after listening to the speech samples, a group of three judges were to write down whatever they heard. The result showed that total cerebral palsic subjects were divided into two subgroups according to their intelligibility according to three speaking conditions. Some subjects showed that speech intelligibility increased greatly if asked to speak 'louder and more accurately'. and the others showed no difference of intelligibility according to the speaking conditions. This study suggested that it would be useful clinically to find out the best instruction to improve intelligibility suitable for each speaker with cerebral palsy.

• Proceedings of the KIEE Conference
• /
• 1998.07b
• /
• pp.654-656
• /
• 1998

This paper deal with the implementation of Active Noise Control (ANC) in a short duct. In case of ANC in the air duct, input microphone, control speaker, error microphone are used. But we can't use input microphone because of the characteristics of short duct. It is difficult to avoid howl. So we propose single-channel adaptive feedback ANC which is composed only error microphone and control speaker without input microphone. FXLMS algorithm is used to compensate for the time delay of the error path. Experimental results show that the controller reduce noise signal sufficiently.