• The Transactions of the Korea Information Processing Society
• /
• v.4 no.8
• /
• pp.2025-2037
• /
• 1997

The concept-based translation system, which is a part of the Korean spoken language translation system, translates spoken utterances from Korean speech recognizer into one of English, Japanese and Korean in a travel planning task. Our system regulates semantic rather than the syntactic category in order to process the spontaneous speech which tends to be regarded as the one ungrammatical and subject to recognition errors. Utterances are parsed into concept structures, and the generation module produces the sentence of the specified target language. We have developed a token-separator using base-words and an automobile grammar corrector for Korean processing. We have also developed postprocessors for each target language in order to improve the readability of the generation results.

• The Journal of the Acoustical Society of Korea
• /
• v.15 no.4
• /
• pp.24-31
• /
• 1996

In this paper, we present a speech translation system for hotel reservation, KT_STS(Korea Telecom Speech Translation System). KT-STS is a speech-to-speech translation system which translates a spoken utterance in Korean into one in Japanese. The system has been designed around the task of hotel reservation(dialogues between a Korean customer and a hotel reservation de나 in Japan). It consists of a Korean speech recognition system, a Korean-to-Japanese machine translation system and a korean speech synthesis system. The Korean speech recognition system is an HMM(Hidden Markov model)-based speaker-independent, continuous speech recognizer which can recognize about 300 word vocabularies. Bigram language model is used as a forward language model and dependency grammar is used for a backward language model. For machine translation, we use dependency grammar and direct transfer method. And Korean speech synthesizer uses the demiphones as a synthesis unit and the method of periodic waveform analysis and reallocation. KT-STS runs in nearly real time on the SPARC20 workstation with one TMS320C30 DSP board. We have achieved the word recognition rate of 94. 68% and the sentence recognition rate of 82.42% after the speech recognition tests. On Korean-to-Japanese translation tests, we achieved translation success rate of 100%. We had an international joint experiment in which our system was connected with another system developed by KDD in Japan using the leased line.

• Phonetics and Speech Sciences
• /
• v.15 no.1
• /
• pp.35-45
• /
• 2023

Tohoku Japanese is known to have voiced stops without pre-voicing in word-initial position, whereas traditional or conservative Japanese has voiced stops with pre-voicing in the same position. One problem with this devoicing of voiced stops is that it affects the distinction between voiced and voiceless stops because their voice onset time (VOT) values overlap. Previous studies have confirmed that Tohoku speakers use post-stop fundamental frequency (F0) as an acoustic cue along with VOT to avoid overlap. However, the role of post-stop F0 as a perceptual cue in this region has barely been investigated. Therefore, this study explored the role of post-stop F0 in stop voicing perception along with VOT. Several perception tests were conducted using resynthesized stimuli, which were manipulated along a VOT continuum orthogonal to an F0 continuum. The results showed no significant regional difference (Tohoku vs. Chubu) for nonsense words (/ta-da/). However, for meaningful words (/pari/ 'Paris' vs. /bari/ 'Bali,' /piza/ 'pizza' vs. /biza/ 'visa'), a significant word effect was found, and it was confirmed that some listeners utilized the post-stop F0 more consistently and steadily than others. Based on these results, we discuss innovative listeners who may lead the change in the perception of stop voicing.

• The Journal of the Acoustical Society of Korea
• /
• v.28 no.2
• /
• pp.155-163
• /
• 2009

In text-to-speech systems, the conversion of text into prosodic parameters is necessarily composed of three steps. These are the placement of prosodic boundaries. the determination of segmental durations, and the specification of fundamental frequency contours. Prosodic boundaries. as the most important and basic parameter. affect the estimation of durations and fundamental frequency. Break prediction is an important step in text-to-speech systems as break indices (BIs) have a great influence on how to correctly represent prosodic phrase boundaries, However. an accurate prediction is difficult since BIs are often chosen according to the meaning of a sentence or the reading style of the speaker. In Japanese, the prediction of an accentual phrase boundary (APB) and major phrase boundary (MPB) is particularly difficult. Thus, this paper presents a method to complement the prediction errors of an APB and MPB. First, we define a subtle BI in which it is difficult to decide between an APB and MPB clearly as a variable break (VB), and an explicit BI as a fixed break (FB). The VB is chosen using the classification and regression tree, and multiple prosodic targets in relation to the pith and duration are then generated. Finally. unit-selection is conducted using multiple prosodic targets. In the MOS test result. the original speech scored a 4,99. while proposed method scored a 4.25 and conventional method scored a 4.01. The experimental results show that the proposed method improves the naturalness of synthesized speech.

• Korean Journal of Environment and Ecology
• /
• v.28 no.4
• /
• pp.385-392
• /
• 2014

Quercus gilva Blume is an evergreen-leaved tree that belongs to Facaceae and grows in tropical and warm-temperate regions, such as Japan, Taiwan, China and Korea. This study was carried out to analyze the distribution and growth characteristics of Q. gilva in the native range. In Korea, this species was distributed in Jeju Island between 80 and 350 meter above sea level, and mainly in the south-western regions of the island from 100 to 200 meter above sea level. The average height of Q. gilva was $9.8{\pm}1.9$ meter, with most trees 9 and 12 meter tall. The average diameter at breast height was $22.6{\pm}6.8$ centimeter, with the majority between 20 and 30 centimeter. Several coppice shoots ($3.8{\pm}2.1$ pieces) were formed from stumps and inferred by frequent cuttings for use. Almost Q. gilva individuals(96.2%) were covered with kinds of vines(18 species), and evergreen Trachelospermim asiaticum (Siebold et Zucc.) Nakai was major species. The analysis of damage of vines revealed that the most dead branches were found with 9 to 12 meter height of Q. gilva, and the more vines attached, the larger damage to Q. gilva individuals. A continuous monitoring and investigation would be required to preserve this species in their habitats.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.15 no.4
• /
• pp.323-328
• /
• 1982

Kareius bicoloratus (Basilewsky) is one of the commonly found right-eye flounders and widely distributed in the coastal waters of Korea and Japan. On December 11,1980, the ailthors carried out an experiment to obtain a large number of fertilized eggs from wild adult fish caught by a trawler. The fish were obtained from Maisaka fish market, Shizuoka Prefecture, Japan. The egg is pelagic, spherical in shape and measuring 1.014-1.04 mm in diameter. The yolk as well as the egg capsule is colorless and transparent, and contain no oil globules. The hatching took place in 73 hr 45 min after fertilization at the water temperature $8.0-9.5^{\circ}C$. Newly hatched larvae are 3.09-3.146 mm in total length, with the anus situated in the middle of the body. The marginal fin does not have Pigment cells and myotome number is 17+20=37. Within one day after hatching, the larvae attained 3.77 mm in total length, and there appeared three or four melanophore on yolk sac. When the larvae attained 3.96 mm in total length, melanophores began to appear on the eye ball. Two days after hatching, the larvae attained 4.05 mm in total length, most of yolk material was absorbed, and the caudal fin began to grow at the terminal part of the notochord. When the larvae attained 4.21 mm in total length, mouth and eyes began to move. After 3 days, the larvae attained 4.342-4.394 mm in total length, alimentary canal differentiated, melanophores appeared on the lower jaw and posterior part of the fin membrane. When the larvae attaind 4.576 mm in total length, marginal line of dorsal fin membrane became concave.