• 한국음향학회:학술대회논문집
• /
• 한국음향학회 1995년도 제12회 음성통신 및 신호처리 워크샵 논문집 (SCAS 12권 1호)
• /
• pp.184-187
• /
• 1995

This paper proposes an algorithm for machine recognition of phonemes in continuous speech. The proposed algorithm is static strategy neural network. The algorithm uses, at the stage of training neuron, features such as PARCOR coefficient and auditory-like perceptual liner prediction . These features are extracted from speech samples selected by a sliding 25.6msec windows with s sliding gap being 3 msec long, then interleaved and summed up to 7 sets of parmeters covering 171 msec worth of speech for use of neural inputs. Perfomances are compared when either PARCOR or auditory-like PLP is included in the feture set.

• 대한음성학회:학술대회논문집
• /
• 대한음성학회 1996년도 10월 학술대회지
• /
• pp.472-472
• /
• 1996

The material of database for speech recognition should include phonetic phenomena as much as possible. At the same time, such material should be phonetically compact with low redundancy[1, 2]. The phonetic phenomena in continuous speech is the key problem in speech recognition. This paper describes the processing of a set of sentences collected from the database of 1993 and 1994 "People's Daily"(Chinese newspaper) which consist of news, politics, economics, arts, sports etc.. In those sentences, both phonetic phenometla and sentence patterns are included. In continuous speech, phonemes always appear in the form of allophones which result in the co-articulary effects. The task of designing a speech database should be concerned with both intra-syllabic and inter-syllabic allophone structures. In our experiments, there are 404 syllables, 415 inter-syllabic diphones, 3050 merged inter-syllabic triphones and 2161 merged final-initial structures in read speech. Statistics on the database from "People's Daily" gives and evaluation to all of the possible phonetic structures. In this sentence set, we first consider the phonetic balances among syllables, inter-syllabic diphones, inter-syllabic triphones and semi-syllables with their junctures. The syllabic balances ensure the intra-syllabic phenomena such as phonemes, initial/final and consonant/vowel. the rest describes the inter-syllabic jucture. The 1560 sentences consist of 96% syllables without tones(the absent syllables are only used in spoken language), 100% inter-syllabic diphones, 67% inter-syllabic triphones(87% of which appears in Peoples' Daily). There are rougWy 17 kinds of sentence patterns which appear in our sentence set. By taking the transitions between syllables into account, the Chinese speech recognition systems have gotten significantly high recognition rates[3, 4]. The following figure shows the process of collecting sentences. [people's Daily Database] -> [segmentation of sentences] -> [segmentation of word group] -> [translate the text in to Pin Yin] -> [statistic phonetic phenomena & select useful paragraph] -> [modify the selected sentences by hand] -> [phonetic compact sentence set]

• 말소리와 음성과학
• /
• 제1권4호
• /
• pp.61-71
• /
• 2009

This study aims to investigate whether the misalignment between syllable and word onsets due to the process of resyllabification affects Korean-English late bilinguals perceiving English continuous speech. Two word-spotting experiments were conducted. In Experiment 1, misalignment conditions (resyllabified conditions) were created by adding CVC contexts at the beginning of vowel-initial words and alignment conditions (non-resyllabified conditions) were made by putting the same CVC contexts at the beginning of consonant-initial words. The results of Experiment 1 showed that detections of targets in alignment conditions were faster and more correct than in misalignment conditions. Experiment 2 was conducted in order to avoid any possibilities that the results of Experiment 1 were due to consonant-initial words being easier to recognize than vowel-initial words. For this reason, all the experimental stimuli of Experiment 2 were vowel-initial words preceded by CVC contexts or CV contexts. Experiment 2 also showed misalignment cost when recognizing words in resyllabified conditions. These results indicate that Korean listeners are influenced by misalignment between syllable and word onsets triggered by a resyllabification process when recognizing words in English connected speech.

• 대한음성학회지:말소리
• /
• 제68권
• /
• pp.115-126
• /
• 2008

In this paper, we propose a vocabulary coverage improvement method for embedded continuous speech recognition (CSR) using knowledgebase. A vocabulary in CSR is normally derived from a word frequency list. Therefore, the vocabulary coverage is dependent on a corpus. In the previous research, we presented an improved way of vocabulary generation using part-of-speech (POS) tagged corpus. We analyzed all words paired with 101 among 152 POS tags and decided on a set of words which have to be included in vocabularies of any size. However, for the other 51 POS tags (e.g. nouns, verbs), the vocabulary inclusion of words paired with such POS tags are still based on word frequency counted on a corpus. In this paper, we propose a corpus independent word inclusion method for noun-, verb-, and named entity(NE)-related POS tags using knowledgebase. For noun-related POS tags, we generate synonym groups and analyze their relative importance using Google search. Then, we categorize verbs by lemma and analyze relative importance of each lemma from a pre-analyzed statistic for verbs. We determine the inclusion order of NEs through Google search. The proposed method shows better coverage for the test short message service (SMS) text corpus.

• 말소리와 음성과학
• /
• 제2권1호
• /
• pp.59-68
• /
• 2010

This study addresses the issue of whether Korean(L1)-English(L2) non-proficient bilinguals are affected by the native coda-neutralization process when recognizing words in English continuous speech. Korean phonological rules require that if liaison occurs between 'words', then coda-neutralization process must come before the liaison process, which results in liaison-consonants being coda-neutralized ones such as /b/, /d/, or /g/, rather than non-neutralized ones like /p/, /t/, /k/, /$t{\int}$/, /$d_{\Im}$/, or /s/. Consequently, if Korean listeners apply their native coda-neutralization rules to English speech input, word detection will be easier when coda-neutralized consonants precede target words than when non-neutralized ones do. Word-spotting and word-monitoring tasks were used in Experiment 1 and 2, respectively. In both experiments, listeners detected words faster and more accurately when vowel-initial target words were preceded by coda-neutralized consonants than when preceded by coda non-neutralized ones. The results show that Korean listeners exploit their native phonological process when processing English, irrespective of whether the native process is appropriate or not.

• 한국음향학회지
• /
• 제20권6호
• /
• pp.46-50
• /
• 2001

본 논문은 HMM (Hidden Markov Model) 음성 인식 시스템에 적용할 수 있는 새로운 인식 시간 알고리즘인 스코아 캐쉬기법을 제안한다. 다른 많은 기법들이 인식 시간을 줄이면서 계산량을 줄이기 위하여 어느 정도의 인식율 저하를 감수하는 반면에 제안하는 스코아 캐쉬기법은 인식율 저하를 전혀 일으키지 않으면서 인식 시간을 상당량 줄일 수 있는 기법이다. 단독어 인식 시스템에 적용 가능할 뿐 아니라 연속어 인식에도 적용이 가능하며, 기존에 이미 설계된 인식 시스템의 구조를 전혀 흩트리지 않고 간단히 하나의 함수만 대치함으로서 인식시간을 크게 감축할 수 있다 또한 기존의 계산량 감축 알고리즘과 함께 적용 가능하므로 추가의 계산량 감소를 얻을 수 있다. 스코아 캐쉬 기법을 적용한 결과 최대 54% 만큼 계산량을 줄일 수 있었다.

• The Journal of the Acoustical Society of Korea
• /
• 제21권1E호
• /
• pp.3-11
• /
• 2002

Broadcast news transcription is one of the hardest tasks in speech recognition because broadcast speech signals have much variability in speech quality, channel and background conditions. We developed a Korean broadcast news speech recognizer. We used a morpheme-based dictionary and a language model to reduce the out-of·vocabulary (OOV) rate. We concatenated the original morpheme pairs of short length or high frequency in order to reduce insertion and deletion errors due to short morphemes. We used a lexicon with multiple pronunciations to reflect inter-morpheme pronunciation variations without severe modification of the search tree. By using the merged morpheme as recognition units, we achieved the OOV rate of 1.7% comparable to European languages with 64k vocabulary. We implemented a hidden Markov model-based recognizer with vocal tract length normalization and online speaker adaptation by maximum likelihood linear regression. Experimental results showed that the recognizer yielded 21.8% morpheme error rate for anchor speech and 31.6% for mostly noisy reporter speech.

• 한국음향학회지
• /
• 제18권1호
• /
• pp.53-60
• /
• 1999

이 논문은 한국 방송 음성 인식에 관한 연구이다. 여기서 우리는 대규모 어휘를 갖는 연속 음성 인식을 위한 방법을 제시한다. 주요 관점은 언어 모델과 탐색 방법이다. 사용된 음성 모델은 기본음소 Semi-continuous HMM이고 언어 모델은 N-gram 방법이다. 탐색 방법은 음성과 언어 정보를 최대한 활용하기 위해 3단계의 방법을 사용하였다. 첫째로, 단어의 끝 부분과 그에 관련된 정보를 만들기 위한 순방향 Viterbi Beam탐색을 하였으며, 둘째로 단어 의 시작 부분과 그에 관련된 정보를 만드는 역방향 Viterbi Beam탐색, 그리고 마지막으로 이들 두 결과와 확률적인 언어 모델을 결합하여 최종 인식결과를 얻기 위해 A/sup */ 탐색을 한다. 이 방법을 사용하여 12,000개의 단어에 대한 화자 독립으로 최고 96.0%의 단어 인식률과 99.2%의 음절 인식률을 얻었다.

• 전기전자학회논문지
• /
• 제11권2호
• /
• pp.89-94
• /
• 2007

본 논문은 연속 밀도 함수를 갖는 HMM별 한국어 연속 음성인식에 관한 연구이다. 여기서 우리는 밀도 함수가 2개에서 44개까지 갖는 연속 HMM모델에서 가장 효율적인 연속 음성인식을 위한 방법을 제시한다. 음성 모델은 36개로 구성한 기본음소를 사용한 CI-Model과 3,000개로 구성한 확장음소를 사용한 CD-Model을 사용하였고, 언어 모델은 N-gram을 이용하여 처리하였다. 이 방법을 사용하여 500개의 문장과 6,486개의 단어에 대하여 화자 독립으로 CI Model에서 최고 94.4%의 단어인식률과 64.6%의 문장인식률을 얻었고, CD Model에서는98.2%의 단어인식률과 73.6%의 문장인식률을 안정적으로 얻었다.

• 대한음성학회지:말소리
• /
• 제62호
• /
• pp.133-147
• /
• 2007

In this paper, we introduce a phone vector discrete HMM(PVDHMM) that decodes a phone sequence string, and demonstrates the applicability to spoken document retrieval. The PVDHMM treats a phone recognizer or large vocabulary continuous speech recognizer (LVCSR) as a vector quantizer whose codebook size is equal to the size of its phone set. We apply the PVDHMM to decode the phone sequence strings and compare the outputs with those of a continuous speech recognizer(CSR). Also we carry out spoken document retrieval experiment through PVDHMM word spotter on the phone sequence strings which are generated by phone recognizer or LVCSR and compare its results with those of retrieval through the phone-based vector space model.