OZEKI KAZUHIKO
| Emeritus Professor etc. | Emeritus Professor |
Researcher Information
Research Keyword
Educational Background
Research Activity Information
MISC
- 重複帯域分割法を用いた雑音に頑健なマルチSNR音声認識
2007, 日本音響学会, 1-P-11 - 確率的な手法による日本語文簡約
2007, 言語処理学会, D5-2 - Sentence compression using statistical information about dependency path length
Kiwamu Yamagata; Satoshi Fukutomi; Kazuyuki Takagi; Kazuhiko Ozeki
This paper is concerned with the use of statistical information about dependency path length for sentence compression. The sentence compression method employed here requires a quantity called inter-phrase dependency strength. In the training process, original sentences are parsed, and the number of tokens is counted for each pair of phrases, connected with each other by a dependency path of certain length, that survive as a modifier-modified phrase pair in the corresponding compressed sentence in the training corpus. The statistics is exploited to estimate the inter-phrase dependency strength required in the sentence compression process. Results of subjective evaluation shows that the present method outperforms the conventional one of the same framework where the distribution of dependency distance is used to estimate the inter-phrase dependency strength., SPRINGER-VERLAG BERLIN, 2006, TEXT, SPEECH AND DIALOGUE, PROCEEDINGS, 4188, 127-134, English, 0302-9743, WOS:000241103500016 - Dependency analysis of spontaneous monologue speech using pause and F0 information: A preliminary study
2006, PS5-20 - クロスVQ歪によるテキスト独立型話者照合における特徴量変換の効果
2006, 日本音響学会講演論文集, 3-2-1, 73-74 - 係り受け経路長を利用した新聞記事の自動簡約
2006, 言語処理学会第12回年次大会発表論文集, P2-11, 236-239 - 文節抽出型文簡約における読みやすさ向上のための文節末修正
2006, 言語処理学会第12回年次大会発表論文集, C5-5, 1087-1090 - VQによるテキスト独立型話者照合における特徴量変換
2006, 日本音響学会講演論文集, 1-11-18, 57-58 - 韻律情報を用いた話し言葉コーパスの係り受け解析の試み
2006, 日本音響学会講演論文集, 3-1-11, 1211-1212 - Japanese dependency structure analysis using information about multiple pauses and F0
LU Meirong; TAKAGI Kazuyuki; OZEKI Kazuhiko
2006, IEICE Trans. INF.& SYST., E89-D, 1, 298-304, 0916-8532 - Noise reduction in time domain using referential reconstruction
IHARA Takehiro; NAGAI Takayuki; OZEKI Kazuhiko; KUREMATSU Akira
2006, IEICE Trans. INF.& SYST., E89-D, 3, 1203-1213, 0916-8532 - 単独発声母音を用いた話者照合における特徴量変換
2006, 日本音響学会誌, 62, 12, 848-855 - 雑音低減のための複素周波数領域における参照再構成法
2006, 電子情報通信学会, NLC2006-40, SP2006-96, 65-70 - 発話様式の違いと基本周波数の関係
2006, 日本法科学技術学会, D-1 - Sentence compression using statistical information about dependency path length
Kiwamu Yamagata; Satoshi Fukutomi; Kazuyuki Takagi; Kazuhiko Ozeki
This paper is concerned with the use of statistical information about dependency path length for sentence compression. The sentence compression method employed here requires a quantity called inter-phrase dependency strength. In the training process, original sentences are parsed, and the number of tokens is counted for each pair of phrases, connected with each other by a dependency path of certain length, that survive as a modifier-modified phrase pair in the corresponding compressed sentence in the training corpus. The statistics is exploited to estimate the inter-phrase dependency strength required in the sentence compression process. Results of subjective evaluation shows that the present method outperforms the conventional one of the same framework where the distribution of dependency distance is used to estimate the inter-phrase dependency strength., SPRINGER-VERLAG BERLIN, 2006, TEXT, SPEECH AND DIALOGUE, PROCEEDINGS, 4188, 127-134, English, 0302-9743, WOS:000241103500016 - Dependency analysis of spontaneous monologue speech using pause and F0 information: A preliminary study
2006, PS5-20 - Japanese dependency structure analysis using information about multiple pauses and F0
2006, IEICE Trans. INF.& SYST., E89-D, 1, 298-304, 0916-8532 - Noise reduction in time domain using referential reconstruction
2006, IEICE Trans. INF.& SYST., E89-D, 3, 1203-1213, 0916-8532 - Feature parameter transformation in speaker verification using vowels uttered in isolation
2006, Journal of the Acoustical Society of Japan, 62, 12, 848-855 - 母音による話者認識における特徴量変換
2005, 電子情報通信学会技術報告, HIP2005-96, 91-96 - マルチSNR・マルチバンド音声認識のためのHMM学習用雑音に関する検討
2005, 日本音響学会講演論文集, 1-P-34, 219-220 - 話者照合におけるGMMの学習とデータ量の関係
2005, 日本音響学会講演論文集, 3-7-12, 125-126 - 参照再構成法を用いた時間周波数マスクによる音声と音楽の分離
2005, 電子情報通信学会, D-SP-4 - マルチSNR・マルチバンド法を用いた話者識別における様々な学習雑音に対する性能評価
2005, 日本音響学会講演論文集, 2-Q-10 - 話し言葉コーパスの係り受け解析を目的とした韻律の分析
2005, 日本音響学会講演論文集, 3-P-29 - 概念距離と係り受けを利用した要約文の文節対応付け
2005, 言語処理学会第11回年次大会発表論文集, D3-8 - 概念距離と係り受けを利用した要約文の文節対応付け
2005, 情報処理学会第67回全国大会講演論文集, 2, 5J-2 - [若手研究者育成レクチャーシリーズ]韻律と統語構造
2004, 電子情報通信学会技術研究報告, 104, 221, 13-18 - マルチパス方式を用いた雑音環境下での単語音声認識 - アクセント情報の利用
2004, 日本音響学会2004年春季研究発表会講演論文集, 3-Q-9, 161-162 - 着目文節の前後のポーズ情報を利用した係り受け解析
2004, 日本音響学会2004年春季研究発表会講演論文集, 2-p-26, 381-382 - 係り受け解析における韻律情報有効性の多数話者による評価
2004, 日本音響学会2004年春季研究発表会講演論文集, 1-7-14, 239-240 - Dependency analysis of read Japanese sentences using pause information: A speaker independent case
2004 - Recovery of dependency structure using multiple pause information
2004 - 重要文抽出と文簡約を併用した新聞記事の自動要約
2004, 言語処理学会第１０回年次大会発表論文集, 4-4 - The use of prosodic information in dependency structure analysis of read Japanese sentences
2004, City University of Hong Kong, 6 - The use of prosodic information in dependency analysis of read Japanese sentences
2004, 4th China-Japan Joint Conference to Promote Cooperation in Natural Language Processing (CJNLP) - [若手研究者育成レクチャーシリーズ]韻律と統語構造
2004, 電子情報通信学会技術研究報告, 104, 221, 13-18 - Automatic adjustment of subband likelihood recombination weights for improving noise-robustness of a multi-SNR multi-band speaker identification system
2004, IEICE Trans. INF&SYST., E87-D, 11, 2453-2459 - Dependency analysis of read Japanese sentences using pause and F0 information: a speaker independent case
2004, 4, 3021-3024 - Improved model training and automatic weight adjustment for multi-SNR multi-band speaker identification system
2004, 3, 1749-1752 - 1/f雑音を用いたマルチSNR部分帯域法による雑音下話者識別
2004, 日本音響学会2004年春季研究発表会講演論文集, [3-Q-8] 159-160 - Dependency analysis of read Japanese sentences using pause information: A speaker independent case
2004 - Recovery of dependency structure using multiple pause information
2004 - The use of prosodic information in dependency structure analysis of read Japanese sentences
2004, City University of Hong Kong, 6 - The use of prosodic information in dependency analysis of read Japanese sentences
2004, 4th China-Japan Joint Conference to Promote Cooperation in Natural Language Processing (CJNLP) - Automatic adjustment of subband likelihood recombination weights for improving noise-robustness of a multi-SNR multi-band speaker identification system
2004, IEICE Trans. INF&SYST., E87-D, 11, 2453-2459 - Dependency analysis of read Japanese sentences using pause and F0 information: a speaker independent case
2004, 4, 3021-3024 - Improved model training and automatic weight adjustment for multi-SNR multi-band speaker identification system
2004, 3, 1749-1752 - 雑音下話者識別におけるマルチSNR 部分帯域法とスペクトルサブトラクション法の性能比較
2003, 日本音響学会講演論文集, 145-146 - Compressing Japanese sentences based on phrase significance and inter-phrase dependency
2003, Proceedings of the 3rd Japan China Natural Language Processing Joint Research Promotion Conference - A neural network approach to dependency analysis of Japanese sentences using prosodic information
2003, 3177-3180 - The use of multiple pause information in dependency structure analysis of spoken Japanese sentences.
2003, 3173-3176 - 日本語読み上げ文の係り受け解析における複数ポーズ情報の利用
2003, 日本音響学会講演論文集, 2-Q-1, 375-376 - マルチSNR部分帯域モデルを用いた話者識別システムの耐雑音性能改善
2003, 日本音響学会講演論文集, 1-4-25, 55-56 - 韻律情報を用いた日本語読み上げ文の係り受け解析におけるニューラルネットワークの利用
2003, 日本音響学会講演論文集, 2-Q-6, 385-386 - サポートベクターマシンによる日本語長文の短文分割
2003, 言語処理学会 第９回年次大会 発表論文集, 453-456 - 係り受け整合度と文節重要度を用いた自動簡約文の主観評価
2003, 言語処理学会 第９回年次大会 発表論文集, 667-670 - The use of prosody for disambiguating Japanese dependency structure
2003, Symposium on Prosody and Speech Processing (Organized by Scientific Research of Priority Areas "Realization of advancedspoken language information processing from prosodic features," Ministry of Science, Culture, Sports, Edulation, Japanese Government) - マルチSNR部分帯域モデルを用いた雑音環境下での話者識別
2003, 日本音響学会誌, 59, 1, 3-12 - Comparison of noise-robustness between multi-SNR subband method and spectral subtraction method in speaker identification under noisy environments
2003, Proceedings of the 2003 fall meeting of the acoustical society of Japan, 145-146 - Compressing Japanese sentences based on phrase significance and inter-phrase dependency
2003, Proceedings of the 3rd Japan China Natural Language Processing Joint Research Promotion Conference - A neural network approach to dependency analysis of Japanese sentences using prosodic information
2003, 3177-3180 - The use of multiple pause information in dependency structure analysis of spoken Japanese sentences.
2003, 3173-3176 - The use of multiple pause information in dependency analysis of Japanese sentences
2003, Proceedings of the 2003 spring meeting of the acoustical society of Japan, 2-Q-1, 375-376 - Improvement of noise-robustness for a speaker identification system using multi-SNR subband model
2003, Proceedings of the 2003 spring meeting of the acoustical society of Japan, 1-4-25, 55-56 - A neural network approach to dependency analysis of Japanese sentences using prosodic information
2003, Proceedings of the 2003 spring meeting of the acoustical society of Japan, 2-Q-6, 385-386 - The use of prosody for disambiguating Japanese dependency structure
2003, Symposium on Prosody and Speech Processing (Organized by Scientific Research of Priority Areas "Realization of advancedspoken language information processing from prosodic features," Ministry of Science, Culture, Sports, Edulation, Japanese Government) - Speaker identification using multi-SNR subband model under noisy environments.
2003, The Journal of the Acoustical Society of Japan, 59, 1, 3-12 - 日本語読み上げ文の係り受け解析における複数ポーズ情報の利用
2002, Proceedings of 2002 2nd Plenary Meeting(organized by Scientific Research of Priority Areas "Realization of advancedspoken language infromation processing from prosodec features," Ministry of Science,Culture,Sports,Education,Japanese Gavernment - Recovery of Japanese dependency structure using prosodic information
Proceedings for 2001 2nd Plenary Meeting and Symposium on Prosody and Speech Processing, Scientific Research of Priority Areas(B), "Realization of advaced spoken language information processing from prosodic features," Ministry of Science, Culture, Sports, Education, 2002, Proceedings for 2001 2nd Plenary Meeting and Symposium on Prosody and Speech Processing, Scientific Research of Priority Areas(B), "Realization of advaced spoken language information processing from prosodic features," Ministry of Science, Culture, Spo・・・, 169-174 - Evaluation of a Japanese sentence compression method based on phrase significance and inter-phrase dependency
2002, 27-32 - Combination of pause and F0 information in dependency anlysis of Japanese sentences
2002, 2, 1173-1176 - 韻律を利用した係り受け解析におけるポーズ・基本周波数情報の結合法の検討
2002, 日本音響学会講演論文集, I, 1-P-22, 1-P-22, 395-396 - Effectiveness of word string language models on noisy broadcast news speech recognition
2002, IEICE Trans. INF.& SYST, E85-D, 7, 1130-1137 - Recovery of Japanese dependency structure using prosodic information
Proceedings for 2001 2nd Plenary Meeting and Symposium on Prosody and Speech Processing, Scientific Research of Priority Areas(B), "Realization of advaced spoken language information processing from prosodic features," Ministry of Science, Culture, Sports, Education, 2002, Proceedings for 2001 2nd Plenary Meeting and Symposium on Prosody and Speech Processing, Scientific Research of Priority Areas(B), "Realization of advaced spoken language information processing from prosodic features," Ministry of Science, Culture, Spo・・・, 169-174 - Evaluation of a Japanese sentence compression method based on phrase significance and inter-phrase dependency
2002, 27-32 - Combination of pause and F0 information in dependency anlysis of Japanese sentences
2002, 2, 1173-1176 - Association method of pause and pitch information in dependency analysis of Japanese sentences
2002, Proceedings of the 2002 spring meeting of the acoustical society of Japan, 1-P-22, 395-396 - Effectiveness of word string language models on noisy broadcast news speech recognition
2002, IEICE Trans. INF.& SYST, E85-D, 7, 1130-1137 - 係り受け解析におけるポーズ・ピッチの利用法の検討
2001, 日本音響学会講演論文集 I, 2-2-8, 2-2-8, 271-272 - The use of prosody in Japanese dependency structure analysis
2001, 123-126 - A multi-SNR subband model for speaker identification under noisy environments
2001, 4, 2849-2852 - Pause information for dependency analysis of read Japanese sentences
2001, 2, 1041-1044 - 総ペナルティ最小化係り受け解析法による音声言語処理
2001, 科学研究費補助金 基盤研究（C)(2)、研究成果報告書（平成９年度）、＜研究代表者：尾関和彦＞ - 文節重要度と係り受け整合度に基づいた文簡約実験
2001, 信学技報, NLC 2001-3, NLC2001-3, 15-20 - Effectiveness of Prosodic Features in Dependency Analysis of Read Japanese Sentences
HIROSE YUKIYOSHI; OZEKI KAZUHIKO; TAKAGI KAZUYUKI
Prosody contains information that is lost when utterances are transcribed into letters or characters. Such information may be useful for syntactic analysis of spoken sentences. In our previous work, we took up 12 prosodic features, and made a statistical model to represent the relationship between those features and dependency distances. Then, using a dependency analyzer that incorporates the model, we have shown that prosodic information is in fact effective for dependency analysis of read Japanese sentences. In the present work, we employed 24 features including new ones, and conducted an extensive search for effective ones. Also, the statistical model was modified to better fit the actual distributions of the feature values. As a result, in open experiments using the ATR 503-sentence database, the correct parsing rate was improved by 21.2% with the use of the prosodic features. This figure is 4.0 points higher than the improvement in the previous experiment of our group. Among the features, the duration of pause was definitely effective in both the open and the closed experiments, while the effectiveness of other features related to the pitch, the power, and the speaking rate, when used together with the duration of pause, was not clear in the open experiments., The Association for Natural Language Processing, 2001, Journal of Natural Language Processing, 8, 4, 71-89, Japanese, 1340-7619, 10011463612, AN10472659 - A Japanese Sentence Compaction Algorithm Based on Phrase Significance and Inter-Phrase Dependency
OGURO REI; OZEKI KAZUHIKO; ZHANG YUJIE; TAKAGI KAZUYUKI
Conventional methods for text summarization are mostly based on the idea of selecting important sentences from a set of given sentences such as a paragraph or a whole text. Those methods have a merit that each selected sentence remains unchanged and is thus correct. However, it is sometimes necessary to shorten each sentence, when a higher compaction rate is required, or when a paragraph-by-paragraph summarization is not adequate. In such sentence compaction, it is important that a shortened sentence is natural as a Japanese sentence. In this paper, the sentence compaction problem is formulated as "a problem of selecting a subsequence of phrases from a given sentence that maximizes the sum of phrase significance scores and inter-phrase dependency scores." Then, an efficient algorithm to solve this problem is proposed. Since this method takes inter-phrase dependency into account, a shortened sentence is expected to be grammatically correct and natural. This paper is focused on the derivation, computational complexity, and implementation issues of the algorithm, and will not discuss the matter of how to define the phrase significance score and the inter-phrase dependency score, though it will be a crucially important matter in practical applications., The Association for Natural Language Processing, 2001, Journal of Natural Language Processing, 8, 3, 3-18, Japanese, 1340-7619, 10011861675, AN10472659 - 係り受け解析における韻律情報の利用
2001, 文部省科学研究費補助金特定領域研究（B）「韻律に着目した音声言語処理の高度化」研究成果報告書（平成12年度）（領域代表者 広瀬啓吉） - フレーム単位で最適SNR部分帯域モデルを選択する話者認識
2001, 日本音響学会講演論文集Ⅰ，3-P-8, 3-P-8, 179-180 - 雑音重畳音声の認識における連語言語モデルの比較
2001, 日本音響学会講演論文集Ⅰ，2-33-18, 2-3-18, 89-90 - Utilization of pause and pitch information for dependency analysis of Japanese sentences
2001, Proceedings of the 2001 fall meeting of the acoustical society of Japan, 2-2-8, 271-272 - The use of prosody in Japanese dependency structure analysis
2001, 123-126 - A multi-SNR subband model for speaker identification under noisy environments
2001, 4, 2849-2852 - Pause information for dependency analysis of read Japanese sentences
2001, 2, 1041-1044 - A Sentence compaction experiment based on phrase significance and inter-phrase dependency
2001, Technical Report of IEICE, NLC2001-3, 15-20 - Effectiveness of prosodic features in dependency analysis of read Japanese sentences
2001, The Journal of Natural Language Processing, 8, 4, 71-89 - A Japanese sentence compaction algorithm based on phrase significance and inter-phrase dependency
2001, The Journal of Natural Language Precessing, 8, 3, 3-18 - Frame-wise selection of optimal SNR subband model for speaker recognition
2001, Proceedings of the 2001 spring meeting of the acoustical society of Japan, 3-P-8, 179-180 - Comparison of word string language models in noisy speech recognition
2001, Proceedings of the 2001 spring meeting of the acoustical society of Japan, 2-3-18, 89-90 - 種々の音響条件におけるニュース音声認識についての考察
2000, 日本音響学会講演論文集Ⅰ，1-Q-10, 1-Q-10, 151-152 - 部分帯域HMMを用いた雑音環境下における話者識別
2000, 日本音響学会講演論文集Ⅰ，1-Q-5, 1-Q-5, 141-142 - 文節重要度と係り受け整合度に基づく文要約アルゴリズム
2000, 言語処理学会第6回年次大会発表論文集, 133-136 - The Application of Decision Trees to Segmentation of Long Japanese Sentences
ZHANG YUJIE; OZEKI KAZUHIKO
It is well known that direct parsing of a long Japanese sentence, including many conjunctive clauses, is extremely difficult. Therefore, it is preferable to segment such a sentence into shorter, simpler ones prior to parsing. Some methods for sentence segmentation have been reported so far. However, because those conventional methods are based on handmade segmentation patterns or rules, they have problems in keeping consistency of the patterns, and in deciding the optimal order of applying those rules. This paper proposes a new method of sentence segmentation using a decision tree, which acquires optimal segmentation patterns and the optimal order of their application automatically from a corpus, taking both linguistic phenomena and their occurrence frequencies into account. Generation and evaluation of a decision tree for sentence segmentation were conducted on an EDR corpus. For 400 evaluation sentences, precision and recall were both 84%, and the percentage of correctly segmented sentences was 77%. It was also confirmed that pruning reduces the tree size significantly without deteriorating the performance., The Association for Natural Language Processing, 2000, Journal of Natural Language Processing, 7, 1, 13-30, Japanese, 1340-7619, 10021991375, AN10472659 - マルチSNR部分帯域GMMを用いた雑音環境下での話者認識
2000, 電子情報通信学会技術研究報告，DSP2000-97,SP2000-63, DSP2000-97, SP2000-63, 55-61 - Effects of word string language models on noisy broadcast news speech recognition
2000, 1, 154-157 - Effectiveness of prosodic features in syntactic analysis of read Japanese sentences
2000, 3, 215-218 - An efficient algorithm for Japanese sentence compaction based on phrase importance and inter-phrase dependency
R Oguro; K Ozeki; YJ Zhang; K Takagi
This paper describes an efficient algorithm for Japanese sentence compaction, First, a measure of grammatical goodness of phrase sequences is defined on the basis of a Japanese dependency grammar. Also a measure of topical importance of phrase sequences is given. Then the problem of sentence compaction is formulated as an optimisation problem of selecting a subsequence of phrases from the original sentence that maximises the sum of the grammatical goodness and the topical importance. A recurrence equation is derived by using the principle of dynamic programming, which is then translated into an algorithm to solve the problem. The algorithm is of polynomial-time with respect to the original sentence length. Finally, an example of sentence compaction is presented., SPRINGER-VERLAG BERLIN, 2000, TEXT, SPEECH AND DIALOGUE, PROCEEDINGS, 1902, 103-108, English, 0302-9743, WOS:000170595900018 - Broadcast news speech recognition under various acoustic conditions
2000, Proceedings of the 2000 spring meeting of the acoustical society of Japan, 1-Q-10, 151-152 - Speaker identification using subband HMMs under noisy environments
2000, Proceedings of the 2000 spring meeting of the acoustical society of Japan, 1-Q-5, 141-142 - The application of decision trees to segmentation of long Japanese sentences
2000, The Journal of Naature Language Processing, 7-1, 13-30 - Speaker recognition using multi-UNR subband GMM under noisy environments
2000, The Institute of Electronics, Information and Communication Engineers Technical Report of IEICE, DSP2000-97, SP2000-63, 55-61 - Effects of word string language models on noisy broadcast news speech recognition
2000, 1, 154-157 - Effectiveness of prosodic features in syntactic analysis of read Japanese sentences
2000, 3, 215-218 - An efficient algorithm for Japanese sentence compaction based on phrase importance and inter-phrase dependency
R Oguro; K Ozeki; YJ Zhang; K Takagi
This paper describes an efficient algorithm for Japanese sentence compaction, First, a measure of grammatical goodness of phrase sequences is defined on the basis of a Japanese dependency grammar. Also a measure of topical importance of phrase sequences is given. Then the problem of sentence compaction is formulated as an optimisation problem of selecting a subsequence of phrases from the original sentence that maximises the sum of the grammatical goodness and the topical importance. A recurrence equation is derived by using the principle of dynamic programming, which is then translated into an algorithm to solve the problem. The algorithm is of polynomial-time with respect to the original sentence length. Finally, an example of sentence compaction is presented., SPRINGER-VERLAG BERLIN, 2000, TEXT, SPEECH AND DIALOGUE, PROCEEDINGS, 1902, 103-108, English, 0302-9743, WOS:000170595900018 - 日本語発話の係り受け解析における韻律情報の利用
1999, 文部省科学研究費補助金 基盤（C) 「韻律に着目した音声言語情報処理の高度化に関する調査研究」 研究成果報告書(平成１０年度）＜研究代表者：広瀬啓吉＞ - Performance comparison of recognition systems : A Bayesian approach
Ozeki Kazuhiko; Ishigami Yoshiyasu; Takagi Kazuyuki
This paper describes a Bayesian approach to performance comparison of recognition systems. Unlike a conventional statistical test, this method makes no decision whether there is a significant difference between the true recognition rate of System A and that of System B. Instead, it gives the probability of the event that the true recognition rate of A is higher than that of B given their recognition results. The probability is referred to as the superiority of A to B. This is similar to a numerical weather forecast, in which what is predicted is the probability of having a certain amount of rain, not a prospect of being sunny or rainy. The superiority is exemplified in various cases for the manner of inputting test data and observing the recognition results, and then its sensitivity for the difference between the respective sample recognition rates of A and B is investigated. All the results support that this method has natural properties which conform to our intuition. The relationship between the superiority in this method and the level of significance in statistical tests is also discussed., Japan Acoustical Society of Japan, 1999, Journal of the Acoustical Society of Japan ?, 20, 3, 171-179, English, 0388-2861, 110003106199, AA00256597 - Speaker identification using subband HMMs
1999, 2, 1019-1022 - 複数の単語bigramモデルを線形結合した言語モデルの検討
1999, 日本音響学会講演論文集Ⅰ，3-Q-20, 3-Q-20, 157-158 - 最小コスト分割問題としての係り受け解析
1999, 言語処理学会第5回年次大会ワークショップ論文集, 9-14 - 単語クラスタリングに基づく言語モデルを用いたニュース音声認識
1999, 日本音響学会講演論文集, I, 1-1-26, 1-1-26, 51-52 - 部分帯域HMMを用いたテキスト独立型話者認識
1999, 日本音響学会講演論文集(]G0001[)，1-1-15, 1-1-5, 29-30 - A classification tree approach to automatic segmentation of Japanese compound sentences
1999, 241-250 - Performance comparison of recognition systems: A Bayesian approach
1999, The Journal of the Acoustical Society of Japan (E), 20, 3, 171-179, 0388-2861 - Speaker identification using subband HMMs
1999, 2, 1019-1022 - A study on a linearly interpolated multiple-bigram language model
1999, Proceedings of the 1999 fall meeting of the acoustical society of Japan, 3-Q-20, 157-158 - News speech recognition using a language model based on word clustering
1999, Proceedings of the 1999 spring meeting of the acoustical society of Japan, 1-1-26, 51-52 - Text-independent speaker identification using subband HMMs
1999, Proceedings of the 1999 spring meeting of the acoustical society of Japan, 1-1-5, 29-30 - A classification tree approach to automatic segmentation of Japanese compound sentences
1999, 241-250 - Performance comparison of recognition systems: a Bayesian approach
1998, 111-116 - 係り受け解析における韻律情報の有効性
1998, 日本音響学会講演論文集，2-Q-26, 2-Q-26, 171-172 - Performance evaluation of word phrase and noun category language models for broadcast news speech recognition
TAKAGI K.
1998, Proc. ICSLP'98, 6, 6, 2507-2510, 10008842683 - The use of prosodic information in syntactic analysis of Japanese utterances
1998, 157-160 - The application of classification trees to bunsetsu segmentation of Japanese sentences
1998, Journal of Natural Language Processing, 5, 4, 17-33 - 分類木を用いた日本語長文の自動分割
1998, 言語処理学会第4回年次大会発表論文集, 390-393 - ニュース音声認識のための言語モデルの比較
1998, 日本音響学会講演論文集，1-6-22, 1-6-22, 47-48 - A tabular method of finding the optimal word string together with its dependency structure
1998, 127-132 - Automatic bunsetsu segmentation of Japanese sentences using a classification tree
1998, 230-235 - Performance comparison of recognition systems: a Bayesian approach
1998, 111-116 - Effectiveness of prosodic information in dependency analysis of Japanese utterances
1998, Proceedings of the 1998 spring meeting of the acoustical society of Japan, 2-Q-26, 171-172 - Performance evaluation of word phrase and noun category language models for broadcast news speech recognition
1998, 6, 2507-2510 - The use of prosodic information in syntactic analysis of Japanese utterances
1998, 157-160 - The application of classification trees to bunsetsu segmentation of Japanese sentences
1998, Journal of Natural Language Processing, 5, 4, 17-33 - Automatin segmentation of long Japanese sentences using a classification tree
1998, Proceedings of the 4th Annual Meetings of the Association for Natural Language Processing, 390-393 - Performance comparison of language models for broadcast news speech recognition
1998, Proceedings of the 1998 spring meeting of the acoustical society of Japan, 1-6-22, 47-48 - A tabular method of finding the optimal word string together with its dependency structure
1998, 127-132 - Automatic bunsetsu segmentation of Japanese sentences using a classification tree
1998, 230-235 - 文節間係り受け距離の統計的性質を用いた日本語文の係り受け解析
1997, 自然言語処理, 4, 2, 3-19 - ニュース音声認識システムの検討
1997, 日本音響学会講演論文集, I, 3-1-9, 1, 103-104 - ニュース音声認識における言語モデルの検討
1997, 情報処理学会・電子情報通信学会研究報告, 97-SLP-19-12, SP97-&NLC97-49, 21-28 - 真の認識率の事後確率に基づく認識システムの性能比較
1997, 電子情報通信学会技術研究報告, SP97-62, 35-42 - 分類木を用いた日本語文の自動文節分割
1997, 情報処理学会自然言語処理研究会資料, 97-NL-121, 1-8 - Syntactic information contained in prosodic features of Japanese utterances
1997, 3, 1471-1474 - Effectiveness of prosodic information in dependency analysis of Japanese sentences
1997, 123-128 - Dependency analysis of Japanese sentences using the statistical property of dependency distance between phrases
1997, The Journal of Natural Language Processing, 4, 2, 3-19 - A study on continuous speech recognition system for broadcast news
1997, Proceedings of the 1997 fall meeting of the acoustical society of Japan, 1, 103-104 - Performance evaluation of language models for broadcast news speech recognition
1997, The Transactions of the Institute of Electronics, Information and Communication Engineers, 97-SLP-19-12, SP97-&NLC97-49, 21-28 - Performance comparison of recognition systems based on the a posteriori probability of true recognition rate
1997, The Institute of Electronics, Information and Communication Engineers Technical Report of IEICE, SP97-62, 35-42 - Automatic bunsetsu segmentation of Japanese sentences using a classification tree
ZHANG Yujie; OZEKI Kazuhiko
In conventional bunsetsu segmentation methods of Japanese sentences, segmentation rules have been given manually. This causes difficuities in maintaining the consistency of the rules, and in deciding an efficient order of rule application. This paper proposes a method of automatic bunsetsu segmentation using a Classification tree, where the knowledge about bunsetsu boundaries can be automatically acquired from a corpus without the need of handwork on rule making. It can also adapt quickly to a new task domain and a new system of morpheme classification. Results of experiments on ATR corpus and EDR corpus show the effectiveness of this method., Information Processing Society of Japan (IPSJ), 1997, The Journal of Natural Language Processing, 1997, 97-NL-121, 1-8, Japanese, 110002934579, AN10115061 - Syntactic information contained in prosodic features of Japanese utterances
1997, 3, 1471-1474 - Effectiveness of prosodic information in dependency analysis of Japanese sentences
1997, 123-128 - 文節スポッテイングにおけるスコアの検討
1996, 日本音響学会平成８年春季講演会, 3-5-3, 113-114 - The mutual information as an acoustic matching measure for speech recognition
1996, 973-978 - Affine transformation based feature normalization for speaker-independent speech recognition
1996, 1113-1116 - 韻律情報を利用した係り受け解析
1996, 電子情報通信学会技術報告, SP96-11, 75-81 - アフィン変換を用いた音声特徴量の正規化
1996, 電子情報通信学会技術報告, 2-8-2, 97-98 - 日本語連続音声認識におけるスコア関数の比較
1996, 電子情報通信学会技術報告, SP96-9, 61-67 - Likelihood normalization using an ergodic HMM for continuous speech recognition
K Ozeki
In recent speech recognition technology, the score of a hypothesis is often defined on the basis of HMM likelihood. As is well known, however, direct use of the likelihood as a scoring function causes difficult problems especially when the length of a speech segment varies depending on the hypothesis as in word-spotting, and some kind of normalization is indispensable. In this paper, a new method of likelihood normalization using an ergodic HMM is presented, and its performance is compared with those of conventional ones. The comparison is made fr om three points of view: recognition rate, word-end detection power, and the mean hypothesis length. It is concluded that the proposed method. gives the best overall performance., I E E E, 1996, ICSLP 96 - FOURTH INTERNATIONAL CONFERENCE ON SPOKEN LANGUAGE PROCESSING, PROCEEDINGS, VOLS 1-4, 4, 2301-2304, English, WOS:A1996BJ20B00580 - 韻律情報を利用した係り受け解析
1996, 日本音響学会講演論文集, 1-P-26, 213-214 - A study on scoring methods in phrase spotting
1996, The Journal of the Acoustical Society of Japan, 3-5-3, 113-114 - The mutual information as an acoustic matching measure for speech recognition
1996, 973-978 - Affine transformation based feature normalization for speaker-independent speech recognition
1996, 1113-1116 - Dependency analysis of Japanese sentences using prosodic information
1996, The Journal of the Acoustic Society of Japan, SP96-11, 75-81 - Speech feature parameter normalization using affine transformation
1996, The Institure of Electronics, Information and Communication Engineers, 2-8-2, 97-98 - Performance comparison of scoring functions for continuous Japanese speech recognition
OGURO Rei; KONDO Norio; OZEKI Kazuhiko
In recent speech recognition technology, the score of a hypothesis is frequently defined on the basis of likelihood calculated with an HMM. The direct use of the likelihood for score is, however, ploblematic especially when the length of a speech segment varies depending on the hypothesis as in word spotting. In this work, we defined several likelihood normalization methods and compared their performances through a number of experiments. The results were assessed from two points of view: the behavior of the score and the recognition rate at true word/phrase boundaries. Overall, normalization by a speech probability calculated with an ergodic HMM gave the best scoring function., The Institute of Electronics, Information and Communication Engineers, 1996, The Institute of Electronics, Information and Communication Engineers, 96, SP96-9, 61-67, Japanese, 110003296233, AN10013221 - Likelihood normalization using an ergodic HMM for continuous speech recognition
1996, 4, 2301-2304 - Dependency analysis of Japanese sentences using prosodic information
1996, Proceedings of the 1996 spring meeting of the acoustical society of Japan, 1-P-26, 213-214 - The mutual information as a scoring function for speech recognition
1995, 電子情報通信学会技術研究報告, SP-95-101, 53-60 - 文節間係り受け距離の統計的性質とその係り受け解析への応用
1995, 電子情報通信学会技術研究報告, NLC95-67, 61-68 - The mutual information as a scoring function for speech recognition
1995, The Institute of Electronics, Information and Communication engineers, SP-95-101, 53-60 - Statistical property of distance between modifier and modified phrases and its application to dependency analysis of Japanese sentences
1995, The Institute of Electronics, Information and Communication Engineers, NLC95-67, 61-68 - DEPENDENCY STRUCTURE-ANALYSIS AS COMBINATORIAL OPTIMIZATION
K OZEKI
This paper is concerned with a problem of selecting optimal word sequences from a word lattice: a problem which often appears in automatic recognition of spoken language. Based on the notion of dependency grammar, a measure of favorability for dependency structures is defined. Then the problem is formulated as that of combinatorial optimization to find the dependency structures which maximize the favorability. A set of recurrence equations are derived, which indicate that the problem allows application of the principle of dynamic programming. The equations lead to a two-phase algorithm which finds optimal word sequences and their optimal dependency structures simultaneously in polynomial time. The order of time complexity is O(M2N5), where M is the number of candidate words in each time segment and N is the length of the word lattice. The algorithm can be considered as a twofold extension of the conventional dependency analyzer. A parallel and layered computation structure to implement the algorithm is also presented., ELSEVIER SCIENCE INC, May 1994, INFORMATION SCIENCES, 78, 1-2, 77-99, English, 0020-0255, WOS:A1994NY31600005 - Performance comparison of recognition systems based on the Akaike information criterion
OZEKI K.
1994, Proc. ICSLP 94, 3, S22-24-4, 1335-1338, 10003787407 - Dependency structure analysis as combinatorial optimization
1994, Information Sciences, 78, 77-99 - Performance comparison of recognition systems based on the Akaike Information Criterion
1994, S22-24-4, 1335-1338 - 誤認識関数を最小化する標準パターン学習アルゴリズム
1993, 電子情報通信学会論文誌, J76-A, 4, 580-588 - 区分線形モデルによる音声認識
1993, 日本音響学会講演論文集, 2-Q-19, 185-186 - 区分最適化学習法による音声認識
1993, 日本音響学会講演論文集, 1-8-6, 11-12 - A multi-template learning algorithm based on minimization of recognition error function
1993, The Transactions of the Institute of Electronics, Information and Communication Engineers, J76-A, 4, 580-588 - Speech recognition using a piecewise linear model
1993, Proceedings of the 1993 fall meeting of the acoustical society of Japan, 2-Q-19, 185-186 - A segmental optimization learning algorithm for speech recognition
1993, Proceedings of the 1993 spring meeting of the acoustical society of Japan, 1-8-6, 11-12 - 構文情報と声調情報を用いる中国語連続音声の認識
1992, 日本音響学会講演論文集, 1-P-3, 109-110 - Recognition of chinese continuous speech using syntatic and tonal information
1992, Proceedings of the 1992 spring meeting of the acoustical society of Japan, 1-P-3, 109-110 - A clustering algorithm to minimize recognition error functions
Akio Ando; Kazuhiko Ozeki
Pattern recognition methods in which multiple templates are prepared for each category and which recognize input patterns using those templates usually partition training patterns into clusters and generate templates as centroids of each cluster. However, most clustering algorithms proposed in the literature aim at minimizing the square error distortion measure when each cluster is represented by its centroid. Therefore the result obtained is not necessarily optimal from the viewpoint of discrimination. This paper describes a new clustering algorithm which is a method of obtaining an optimum solution to a template generation problem. The simulated annealing is used for searching for a cluster partition that minimizes the “degree” of recognition error when all training patterns are recognized by using centroids as templates. This paper also describes a vowel template generation problem in speech recognition as an example of pattern discrimination problems. The new algorithm yielded better experimental results than the LBG algorithm and the LVQ algorithm. Copyright © 1991 Wiley Periodicals, Inc., A Wiley Company, 1991, Electronics and Communications in Japan (Part III: Fundamental Electronic Science), 74, 8, 89-98, English, 1520-6440, 0026207853 - 誤認識関数を最小化するクラスタリングアルゴリズム
1991, 電子情報通信学会論文誌, J74-A, 3, 360-367 - 誤認識関数を最小化する標準パターン学習アルゴリズム
1991, 日本音響学会講演論文集, 3-P-29, 205-206 - 単語間意味関係のグラフ理論的解析
1991, 電子情報通信学会技術研究報告, NLC90-52, 25-32 - 誤認識関数を最小化する標準パターン学習アルゴリズム
1991, 421-426 - A clustering algorithm to minimize recognition error functions
Akio Ando; Kazuhiko Ozeki
Pattern recognition methods in which multiple templates are prepared for each category and which recognize input patterns using those templates usually partition training patterns into clusters and generate templates as centroids of each cluster. However, most clustering algorithms proposed in the literature aim at minimizing the square error distortion measure when each cluster is represented by its centroid. Therefore the result obtained is not necessarily optimal from the viewpoint of discrimination. This paper describes a new clustering algorithm which is a method of obtaining an optimum solution to a template generation problem. The simulated annealing is used for searching for a cluster partition that minimizes the “degree” of recognition error when all training patterns are recognized by using centroids as templates. This paper also describes a vowel template generation problem in speech recognition as an example of pattern discrimination problems. The new algorithm yielded better experimental results than the LBG algorithm and the LVQ algorithm. Copyright © 1991 Wiley Periodicals, Inc., A Wiley Company, 1991, Electronics and Communications in Japan (Part III: Fundamental Electronic Science), 74, 8, 89-98, English, 1520-6440, 0026207853 - A clustering algorithm to minimize recognition error function
1991, The Transactions of the Institute of Electronics, Information and Communication Engineers, J74-A, 3, 360-367 - A Multi-template learning alogrithm based on minimization of recognition error function
1991, Proceedings of the 1991 spring meeting of the acoustical society of Japan, 3-P-29, 205-206 - A graph theoretical analysis of semantic relation among words
1991, The Transactions of the Institute of Electronics, Information and Communication Engineers, NLC90-52, 25-32 - A MULTI-TEMPLATE LEARNING ALGORITHM BASED ON MINIMIZATION OF RECOGNITION ERROR FUNCTION
A ANDO; K OZEKI
ELSEVIER SCIENCE PUBL B V, 1991, ARTIFICIAL NEURAL NETWORKS, VOLS 1 AND 2, 421-426, English, WOS:A1991BV08V00068 - 識別率を評価関数とした最適なクラスタリング
1990, 日本音響学会講演論文集, 3-3-4, 77-78 - 音声認識のための確率的言語モデル
1990, AVIRG１２月例会資料 - Clustering algorithms to minimize recognition error function and their applications to the vowel template learning
1990, 7-5-1, 217-220 - Statistical language modelling for speech recognition
1990, 67-80 - A polynomial-order algorithm for optimal phrase sequence selection from a phrase lattice and its parallel layered implementations
1990, 2, 311-316 - An optimization of clustering taking recognition rate into consideration
1990, Proceedings of the 1990 spring meeting of the acoustical society of Japan, 3-3-4, 77-78 - Statistical language modelling for speech recognition
1990, AVIRG (Auditory and Visual Information Research Group) - Clustering algorithms to minimize recognition error function and their applications to the vowel template learning
1990, 7-5-1, 217-220 - Statistical language modelling for speech recognition
1990, 67-80 - A polynomial-order algorithm for optimal phrase sequence selection from a phrase lattice and its parallel layered implementations
1990, 2, 311-316 - 音声認識システムのための子音認識パラメータの検討
1989, 日本音響学会講演論文集, 1-3-5, 9-10 - 音声認識システムのための初期母音標準パターンの構成法
1989, 日本音響学会講演論文集, 2-1-2, 55-56 - 識別という観点から最適にクラスタリングを行う母音標準パターンの学習法
1989, 電子情報通信学会技術研究報告, SP89-77, 9-16 - A Preliminary evaluation of consonant recognition parameters for speech recognition system
1989, Proceedings of the 1989 fall meeting of the acoustical society of Japan, 1-3-5, 9-10 - A preliminary study of initial vowel templates for speech recognition system
1989, Proceedings of the 1989 fall meeting of the acoustical society of Japan, 2-1-2, 55-56 - An optimization of clustering taking recognition rate into consideration and its application to the vowel templates learning
1989, The Transactions of the Institute of Electronics, Information and Communication Engineers - 多文節間の係り受け整合度に基づき最適文節列を選択する多段決定アルゴリズム
1988, 電子情報通信学会論文誌, J71-D, 4, 669-677 - 音声認識のための最適文節列選択アルゴルズム
1988, 電子情報通信学会春季全国大会, SA-1-2, 1-343 - 1-344 - 模擬アニーリング法による母音標準パターンの最適配置
1988, 日本音響学会講演論文集, 2-P-22, 303-304 - A multi-stage decision algorithm for optimum bunsetsu sequence selection based on the degree of multi-bunsetsu kakariuke-dependency
1988, The Transactions of the Institute of Electronics, Information and Communication, J71-D, 4, 669-677 - Algorithms for optimum bunsetsu sequence selection in speech recognition
1988, The Transactions of the Institute of Electronics, Information and Communication Engineers, SA-1-2, 1-343 - 1-344 - Optimal arrangement of vowel templates using simulated annealing
1988, Proceedings of the 1988 fall meeting of the acoustical society of Japan, 2-P-22, 303-304 - 文節ラチスから最適係り受け構造を選択する多段決定アルゴリズム
1987, 電子情報通信学会論文誌, J70-D, 12, 2621-2629 - 係り受けの整合度に基づき最適文節列を選択する多段決定アルゴリズム
1987, 電子情報通信学会論文誌, J70-D, 3, 601-609 - 文節ラチスから最適係り受け構造を選択する多段決定アルゴリズム
1987, 電子情報通信学会論文誌, J70-D, 12 - 母音標準パターンの自動獲得法
1987, 電子通信学会技術研究報告, SP86-88, 7-14 - 母音標準パターン自動獲得法のための学習アルゴリズムの検討
1987, 日本音響学会講演論文集, 1-5-11, 21-22 - 前フレームの分析結果を利用する繰り返し的ホルマント抽出法
1987, 日本音響学会講演論文集, 2-6-7, 161-162 - 係り受けの整合度に基づき最適文節列を選択する一連のアルゴリズムとその計算量の比較
1987, 電子情報通信学会技術研究報告, COMP87-5, 39-48 - 多文節間係り受け整合度を用いたときの最適文節列選択アルゴリズム
1987, 日本音響学会講演論文集, 3-5-8, 127-128 - 標準パターン間分離度による母音標準パターン自動獲得の制御
1987, 日本音響学会講演論文集, 2-5-12, 73-74 - A multi-stage decision algorithm to select optimum kakariuke structures from bunsetsu lattice
1987, The Transactions of the Institute of Electronics, Information and Communication, J70-D, 12, 2621-2629 - A multi-stage decision algorithm to select optimum bunsetsu sequences based on degree of kakariuke-dependency
1987, The Transactions of The Institute of Electronics,Information and Communication, J70-D, 3, 601-609 - A multi-stage decision algorithm to select optimum kakariuke structure from bunsetsu lattice
1987, The Transactions of the Institute of Electronics, Information and Communication Engineers, J70-D, 12 - A method for automatic acquisiton of vowel templates
1987, The Transaction of the Institute of Electronics and Communication Engineers of Japan, SP86-88, 7-14 - A comparison of learning algorithms for automatic aquisition of vowel templates
1987, Proceedings of the 1987 spring meeting of the acoustical society of Japan, 1-5-11, 21-22 - A Formant extraction method using estimated data from previous frame
1987, Proceedings of the 1987 spring meeting of the acoustical society of Japan, 2-6-7, 161-162 - A series of algorithms to select optimum bunsetsu sequence based on the degree of kakariuke-dependency and comparison of their amount of computation
1987, The Transactions of the Institute of Electronics,Information and Communication Engineers, COMP87-5, 39-48 - An optimum bunsetsu sequence selection algorithm based on the degree of multi-bunsetsu kakariuke-dependency
1987, Proceedings of the 1987 fall meeting of the acoustical society of Japan, 3-5-8, 127-128 - Control of a learning algorithm for vowel templates aquisition using vowel templates separability
1987, Proceedings of the 1987 fall meeting of the acoustical society of Japan, 2-5-12, 73-74 - 連続音声認識システムにおける母音標準パターンの自動作成
1986, 日本音響学会講演論文集, 3-1-15, 113-114 - 最適文節列を選択するための多段決定アルゴリズム
1986, 電子通信学会技術研究報告, SP86-32, 41-48 - 母音標準パターンの自動獲得法
1986, 日本音響学会講演論文集, 3-3-7, 103-104 - Maximum likelihood estimation of power-spectrum of speech signals with additive white noise.
1986, The Transactions of the Institute of Electronics and Communication Engineers of Japan, E69, 11, 1153-1156 - 最適文節列を選択するための多段決定アルゴリズムとその制御
1986, 日本音響学会講演論文集, 2-3-1, 53-54 - 文節ラティスから最適文節列を選択するための多段決定アルゴリズム
1986, 電子通信学会技術研究報告, COMP86-47, 47-57 - A `bunsetsu´speech recognition system using a top-down processor controlled by bottom-up information
1986, 21.17.1-21.17.4, 1129-1132 - A preliminary study on automatic generation of vowel templates for continuous speech recognition system
1986, Proceedings of the 1986 spring meeting of the acoustical society of Japan, 3-1-15, 113-114 - A multi-stage decision algorithm for optimum bunsetsu sequence selection
1986, The Transactions of the Institute of Electronics and Communication Engineers of Japan, SP86-32, 41-48 - A method for automatic acquisition of vowel templates
1986, Proceedings of the 1986 fall meeting of the acoustical society of Japan, 3-3-7, 103-104 - Maximum likelihood estimation of power-spectrum of speech signals with additive white noise.
1986, The Transactions of the Institute of Electronics and Communication Engineers of Japan, E69, 11, 1153-1156 - A multi-stage decision algorithm for optimum bunsetsu sequence selection and its control
1986, Proceedings of the 1986 fall meeting of the acoustical society of Japan, 2-3-1, 53-54 - A multi-stage decision algorithm for optimum bunsetsu sequence selection from bunsetsu lattice
1986, The Transactions of the Institute of Electronics and Communication Engineer of Japan, COMP86-47, 47-57 - A `bunsetsu´speech recognition system using a top-down processor controlled by bottom-up information
1986, 21.17.1-21.17.4, 1129-1132 - 文節を入力単位とした連続音声認識システムの検討
1985, 日本音響学会講演論文集, 3-4-6, 91-92 - 文節を入力単位とした連続音声認識システム
1985, 日本音響学会 音声研究会資料, S85-14, 99-106 - 連続音声認識システムにおける無声化母音標準パターン自動作成の試み
1985, 日本音響学会講演論文集, 3-4-15, 123-124 - 連続音声認識システムにおける鼻音化母音標準パターンの自動作成
1985, 電子通信学会 情報・システム部門全国大会, S5-9, 1-343 - 1-344 - 連続音声認識システムにおける母音標準パターンの自動作成
1985, 日本音響学会音声研究会資料, S85-60, 461-468 - Consideration on a continuous speech recognition system using "bunsetsu" as input units
1985, Proceedings of the 1985 spring meeting of the acoustical society of Japan, 3-4-6, 91-92 - A continuous speech recognition system using "Bunsetsu"as input units
1985, The Journal of the Acoustical Society of Japan, S85-14, 99-106 - A preliminary study on automatic production of unvoiced vowel templates for continuous speech recognition system
1985, Proceedings of the 1985 fall meeting of the acoustical society of Japan, 3-4-15, 123-124 - A preliminary study on automatic production of nasalized vowel templates for continuous speech recognition system
1985, The Transactions of the Institute of Electronics and Communication Engineers of Japan, S5-9, 1-343 - 1-344 - A preliminary study on automatic generation of vowel templates for continuous speech recognition system
1985, Transactions of the committee on speech research The acoustical society of Japan, S85-60, 461-468 - 母音間分離度による認識パラメータの評価
1984, 日本音響学会講演論文集, 2-3-10, 151-152 - ボトムアップ情報によるトップダウン処理部の制御を用いた連続音声認識システム
1984, 日本音響学会講演論文集, 2-9-15, 71-72 - 連続音声認識のための音韻認識部の構成
1984, 日本音響学会講演論文集, 2-9-16, 73-74 - An Adaptive filtering algorithm using an orthogonal projection to an affine subspace and its properties
1984, Electronics and Communications in Japan (scripta publishing company, traslated from Denshi Tsushin Gakkai Ronbunshi), 67-A, 5,19-27 - アフィン部分空間への直交射影を用いた適応フィルタ・アルゴリズムとその諸性質
1984, 電子通信学会論文誌, J67-A, 2, 126-132 - Evaluation of speech recognition parameters using vowel separability measure
1984, Proceedings of the 1984 spring meeting os the acoustical society of Japan, 2-3-10, 151-152 - A continuous speech recognition system having a top-down processor controlled by bottom-up information
1984, Proceedings of the 1984 fall meeting of the acoustical society of Japan, 2-9-15, 71-72 - Construction of a phoneme recognizer for continuous speech recognition
1984, Proceedings of the 1984 fall meeting of the acoustical society of Japan, 2-9-16, 73-74 - An Adaptive filtering algorithm using an orthogonal projection to an affine subspace and its properties
1984, Electronics and Communications in Japan (scripta publishing company, traslated from Denshi Tsushin Gakkai Ronbunshi), 67-A, 5,19-27 - An adaptive filtering algorithm using an orthogonal projection to an affine subspace and its properties
1984, The Transactions of the Institute of Electronics and Communication Engineers of02 Japan, J67-A, 2, 126-132 - Suppression of howling between microphones and monitoring speakers -An application of an adaptive filter-
1983, 6, 103-106 - Suppression of howling between microphones and monitoring speakers -An application of an adaptive filter-
1983, 6, 103-106 - APA:アフィン射影を用いた適応アルゴリズム
1982, 電子通信学会 光・電波部門全国大会, S1-23, 1-283,-284 - 高速マイクロ・プロセッサーを用いた雑音除去
1982, 電子通信学会技術研究報告, EA81-58, 39-46 - 適応フィルターを用いたハウリング抑圧装置
1982, 日本音響学会講演論文集, 1-7-9, 61-62 - アフィン射影を用いた適応アルゴルズムの諸性質
1982, 日本音響学会講演論文集, 3-6-10, 153-154 - アフィン射影を用いた適応アルゴルズムの諸性質
1982, 日本音響学会 音声研究会資料, S82-52, 409-416 - Noise cancelling using high-speed micro-processor
1982, The Transactions of the Institute of Electonics and Communication Engineers of Japan, EA81-58, 39-46 - A howling suppressor using an adaptive filter
1982, Proceedings of the 1982 fall meeting of the acoustical society of Japan, 1-7-9, 61-62 - Properties of the adaptation algorithm based on the affine projection
1982, Proceedings of the 1982 fall meeting of the acoustical society of Japan, 3-6-10, 153-154 - Properties of the adaptation algorithm based on the affine projection
1982, The Journal of the Acoustical Society of Japan, S82-52, 409-416 - A rapidly converging algorithm for adaptive noise cancelling filters
1981, 102nd Meeting of the Acoustical Society of America - A rapidly converging algorithm for adaptive noise cancelling filters
1981, 102nd Meeting of the Acoustical Society of America - 適応フィルターの為の新しい係数修正アルゴリズム
1981, 日本音響学界講演論文集, 2-7-10, 505-506 - 適応フィルターを用いた拡声系の音質改善
1981, 日本音響学会講演論文集, 1-2-10, 589-590 - 適応フィルターにおける収束速度向上のための一方法
1981, 日本音響学会 音声研究会資料, S81-10, 73-80 - 高速マイクロコンピュータを用いた雑音除去実験
1981, 日本音響学会講演論文集, 3-2-9, 153-154 - 新アルゴリズムを用いた適応フィルターの収束速度と安定性
1981, 日本音響学会講演論文集, 1-1-14, 28-28 - A rapidly converging algorithm for adaptive noise cancelling filters
1981, 102nd Meeting of the Acoustical Society of America - A rapidly converging algorithm for adaptive noise cancelling filters
1981, 102nd Meeting of the Acoustical Society of America - A new training algorithm for adaptive filters
1981, Proceedings of the 1981 spring meeting of the acoustical society of Japan, 2-7-10, 505-506 - A method of quality improvement for sound-enforcement by using an adaptive filter
1981, Proceedings of the 1981 spring meeting of the acoustical society of Japan, 1-2-10, 589-590 - A method of improving the speed of convergence in adaptive filters
1981, The Journal of the Acoustical Society of Japan, S81-10, 73-80 - Experimental noise cancelling using high speed μ processor
1981, Proceedings of the 1981 fall meeting of the acoustical society of Japan, 3-2-9, 153-154 - Convergence rate and stability of an adaptive filter using a new training algorithm
1981, Proceedings of the 1981 fall meeting of the acoustical society of Japan, 1-1-14, 28-28 - 情報理論を用いた主成分分析理論の再構築
1980, 情報理論とその応用研究会、第三回研究討論会資料, 539-546 - 有色雑音が加わった音声信号からの予測係数の最尤推定
1980, 日本音響学会講演論文集, 2-7-4, 589-590 - 白色雑音が加わった音声信号からの予測係数の最尤推定
1980, 日本音響学会講演論文集, 1-1-6, 369-370 - 超準解析の手法による連続分布のエントロピーの新しい定義
1980, 電子通信学会論文誌, 63-A, 4, 286-287 - A NEW DEFINITION OF THE ENTROPY OF GENERAL PROBABILITY-DISTRIBUTIONS BASED ON THE NONSTANDARD ANALYSIS
K OZEKI
ACADEMIC PRESS INC JNL-COMP SUBSCRIPTIONS, 1980, INFORMATION AND CONTROL, 47, 2, 94-106, English, 0019-9958, 80000950593, WOS:A1980LN21000002 - Maximum likelihood estimation of predictor coefficients from speech signals corrupted by colored noise
1980, Proceedings of the 1980 spring meeting of the acoustical society of Japan, 2-7-4, 589-590 - Maximum likelihood estimation of predictor coefficients from speech signals in the presence of white noise
1980, Proceedings of the 1980 fall meeting of the acoustical society of Japan, 1-1-6, 369-370 - A new definition of the entropy of continuous distributions based on the Non-standard analysis.
1980, The Transactions of the Institute of Electronics and Communication Engineers of Japan, 63-A, 4, 286-287 - A new definition of the entropy of general probability distributions based on the non-standard analysis
1980, Information and Control, 47, 2, 94-106, 0019-9958, 80000950593 - 座標変換に対して不変な形式を持つ統計的データの次元圧縮理論
1979, 視聴覚情報研究会 ９月例会 - 雑音が加わった音声信号からの予測係数の最尤推定
1979, 日本音響学会講演論文集, 2-2-6, 523-524 - 録音編集合成による天気予報文の編集装置
1979, 日本音響学会講演論文集, 1-5-2, 157-158 - 雑音が加わった音声信号からの予測係数の最尤推定
1979, 日本音響学会 音声研究会資料, SP79-44, 333-340 - 座標変換に対して不変な形式を持つ統計的データの次元圧縮理論
1979, 電子通信学会技術研究報告, PRL79-48, 9-16 - 文生成と情報伝達
1979, 電子情報学会技術研究報告, AL78-92, 33-40 - A Coordinate-free theory of eigenvalue analysis related to the method of principal components and the Karhunen-Lo(]J1118[)ve expansion
1979, Information and Control, 42, 1, 38-59, 0019-9958, 80014054252 - Dimenson compression of statistical data A coordinate-free theory
1979, AVIRG - Maximum likelihood estimation of prediction coefficients from speech signals corrupted by noise.
1979, Proceedings of the 1979 fall meeting of the acoustical society of Japan, 2-2-6, 523-524 - An experimental speech synthesis system with pre-recorded words and phrases for local weather report
1979, Proceedings of the 1979 fall meeting of the acoustical society of Japan, 1-5-2, 157-158 - The Journal of the Acoustical Society of Japan
1979, Maximum likelihood estimation of predictor coefficients from noisy speech signals, SP79-44, 333-340 - Dimension compression of statistical data A coordinate-free theory
1979, The Transactions of the Institute of Electronics and Communication Engineers of Japan, PRL79-48, 9-16 - Sentence production and information transmission
1979, The Transactions of the Institute of Electronics and Communication Engineers of Japan, AL78-92, 33-40 - A Coordinate-free theory of eigenvalue analysis related to the method of principal components and the Karhunen-Lo(]J1118[)ve expansion
1979, Information and Control, 42, 1, 38-59, 0019-9958, 80014054252 - 自己回帰パラメータ空間のRiemann幾何学的考察
1977, 日本音響学会音声研究資料, S77-04 - 自己回帰パラメータ空間上のRiemann 計量
1977, 日本音響学会講演論文集, 3-2-21, 389-390 - 観測誤差を考慮に入れたKarhunen-Loeve 展開
1977, 電子通信学会技術研究報告, PRL 77-22, 21-30 - A riemannian metric structure on autoregressive parameter spaces
1977, The Journal of the Acoustical Society of Japan, S77-04 - A riemannian metric on the autoregressive parameter space
1977, Proceedings of the 1977 spring meeting of the acoustical society of Japan, 3-2-21, 389-390 - A modification of the Karhunen-loeve expansion theory by taking observation noise into consideration
1977, The Transactions of the Institute of Electronics and Communication Engineers of Japan, PRL 77-22, 21-30 - 複合時のアクセントにおける規範と実現の関係
1976, 日本音響学会講演論文集, 2-5-19, 213-232 - 男声・女声変換実験
1976, 日本音響学会誌, 32, 6, 362-368 - On the accentual concept and its realization in total connection between words
1976, Proceedings of the 1976 spring meeting of the acoustical society of Japan, 2-5-19, 213-232 - An experiment on male to female voice conversion
1976, The Journal of the Acoustical Society of Japan, 32, 6, 362-368 - 音調制御編集合成実験
1975, 日本音響学会講演論文集, 1-4-4, 107-108 - 男声、女声変換実験
1975, 電子情報学会技術研究報告, EA75-29, 49-56 - Encoding of linguistic information generated by a Markov chain which is associated with a stochastic context-free grammar
1975, System, Computer, Controls (scripta publishing company,translated from Denshi Tsushin Gakkai Ronbunshi), 6-3, 75-80 - 確率的文脈自由形文法に随伴するマルコフ連鎖により生成される言語情報の符号化
1975, 電子通信学会論文誌, 58-D, 6, 322-327 - 確率的文脈自由形言語のエルゴード性とその応用
1975, 電子通信学会論文誌, 58-D, 4, 216-217 - Some trials on the control of tonality in synthesizing speech by parameter edition
1975, Proceedings of the 1975 fall meeting of the acoustical society of Japan, 1-4-4, 107-108 - An experiment on male to female voice conversion
1975, The Transactions of the Institute of Electronics and Communication Engineers of Japan, EA75-29, 49-56 - Encoding of linguistic information generated by a Markov chain which is associated with a stochastic context-free grammar
1975, System, Computer, Controls (scripta publishing company,translated from Denshi Tsushin Gakkai Ronbunshi), 6-3, 75-80 - Encoding of linguistic information generated by a Markov chain which is associated with a stochastic context-free grammar
1975, The Transactions of the Institute of Electronics and Communication Engineers of Japan, 58-D, 6, 322-327 - Ergodicity of stochastic context-free languages and its application
1975, The Transactions of the Institute of Electronics and Communication Engineers of Japan - 確率的文脈自由形文法とマルコフ連鎖
1974, 電子通信学会 オートマトンと言語研究会資料, AL73-76 - 言語情報の最適符号化
1974, 電子通信学会 オートマトンと言語研究会資料, AL73-74 - Optimal encoding of linguistic informantion
1974, Systems, Computers, Controls (scripta publishin company,translated from Denshi Tsushin Gakkai Ronbunshi), 5-3, 96-103 - Relation between channel capacity of a context-free grammar and transmission rate of a stochastic context-free grammar
1974, Systems, Computers, Controls (scripta publishing company, translated from Denshi Tsushin Gakkai Ronbunshi), 5-7, 107-114 - Stochastic context-free grammar and Karkov chain
1974, Systems, Computers, Controls (scripta publishing company,translated from Denshi Tsushin Gakkai Ronbunshi), 5-3, 104-110 - 文脈自由形文法の通信路容量と確率的文脈自由形文法の伝送速度との関係
1974, 電子通信学会論文誌, 57-D, 12, 716-723 - 確率的文脈自由形文法とマルコフ連鎖
1974, 電子通信学会論文誌, 57-D, 6, 369-375 - 言語情報の最適符号化
1974, 電子通信学会論文誌, 57-D, 6, 361-368 - 言語情報の最適符号化
1974, 電子通信学会論文誌, 57-D, 6 - Sur l'espace topologique li(]J1117[) (]J1102[) une nouvelle th(]J1117[)orie de l'apprentissage
1974, Kybernetik, 14, 141-149 - Stochastic context-free grammars and markov chains
1974, The Transactions of the Institute of Electronics and Communication Engineers of Japan, AL73-76 - Optimal encoding of linguistic information
1974, The Transactions of the Insitute of Electronics and Communication Engineers of Japan, AL73-74 - Optimal encoding of linguistic informantion
1974, Systems, Computers, Controls (scripta publishin company,translated from Denshi Tsushin Gakkai Ronbunshi), 5-3, 96-103 - Relation between channel capacity of a context-free grammar and transmission rate of a stochastic context-free grammar
1974, Systems, Computers, Controls (scripta publishing company, translated from Denshi Tsushin Gakkai Ronbunshi), 5-7, 107-114 - Stochastic context-free grammar and Karkov chain
1974, Systems, Computers, Controls (scripta publishing company,translated from Denshi Tsushin Gakkai Ronbunshi), 5-3, 104-110 - Relation between channel capacity of a context-free grammar and transmission rate of a stochastic context-free grammar
1974, The Transactions of the Institute of Electronics and Communication Engineers of Japan, 57-D, 12, 716-723 - Stochastic context-free grammars and Markov chain
1974, The Transactions of the Institute of Electronics and Communication Engineers of Japan, 57-D, 6, 369-375 - Optimal encoding of linguistic information
1974, The Transactions of the Institute of Electronics and Communication Engineers of Japan, 57-D, 6, 361-368 - Optimal encoding of linguistic information
1974, The Transactions of the Institute of Electronics and Communication Engineers of Japan, 57-D, 6 - Sur l'espace topologique li(]J1117[) (]J1102[) une nouvelle th(]J1117[)orie de l'apprentissage
1974, Kybernetik, 14, 141-149 - ある線形オートマトンの同定
1973, 電子通信学会 オートマトンと言語研究会資料, AL72-132 - 確率的文脈自由形文法の標準形
1973, 電子通信学会 オートマトンと言語研究会資料, AL72-130 - Identification of a linear automaton
1973, Sysems, Computers, Controls (scripta publishing company,translated from Denshi Tsushin Gakkai Ronbunshi), 4-5, 57-63 - 学習可能性の理論
1973, 電子情報学会論文誌, 56-D, 7, 416-423 - ある線形オートマトンの同定について
1973, 電子通信学会論文誌, 56-D, 10, 584-590 - 学習空間の位相的性質
1973, 電子通信学会論文誌, 56-D, 10, 561-568 - A Theory of Learnability
1973, Kybernetik, 13, 123-131 - Topological aspects of learning space
1973, Systems, Computers, Controls (scripta publishing company,traslated from Denshi Tsushin Gakkai Ronbunshi), 4-5, 40-47 - Identification of a linear automaton
1973, The Transactions of the Institute of Electronics and Communication Engineers of Japan, AL72-132 - Normal forms of stochastic context-free grammars
1973, The Transactions of the Institute of Electonics and Communication Engineers of Japan, AL72-130 - Identification of a linear automaton
1973, Sysems, Computers, Controls (scripta publishing company,translated from Denshi Tsushin Gakkai Ronbunshi), 4-5, 57-63 - A theory of learnability
1973, The Transactions of the Institute of Electronics and Communication Engineers of Japan, 56-D, 7, 416-423 - Identification of a linear automaton
1973, The Transactions of The Institute of Electronics and Communication Engineers of Japan, 56-D, 10, 584-590 - Topological aspects of learning space
1973, The Transactions of the Institute of Electronics and Communication Engineers of Japan, 56-D, 10, 561-568 - A Theory of Learnability
1973, Kybernetik, 13, 123-131 - Topological aspects of learning space
1973, Systems, Computers, Controls (scripta publishing company,traslated from Denshi Tsushin Gakkai Ronbunshi), 4-5, 40-47 - 文脈自由系言語の情報論的性質
1972, 電子通信学会 オートマトンと言語研究会資料, AL72-55 - 文脈自由形言語の情報論的性質
1972, 電子通信学会論文誌, 55-D, 11, 753-760 - 学習可能性の理論
1972, 電子通信学会論文誌, AL, PRL72-130, - 連想形記憶の二，三の性質
1972, 電子通信学会論文誌, 55-D, 5, 323-330 - Information theoretic aspects of context-free languages
1972, The Trasactions of the Institute of Electonics and Communication Engineers of Japan, AL72-55 - Information theoretic aspects of context-free languages
1972, The Transactions of The Institute of Electronics and Communication Engineers of Japan, 55-D, 11, 753-760 - A theory of learnability
1972, The transactions of the institute of electronics and communication engineers of Japan, AL, PRL72-130, - On some properties of associative memory
1972, The transactions of the institute of electronics and communication engineers of Japan, 55-D, 5, 323-330