記事
Differenced maximum mutual information criterion for robust unsupervised acoustic model adaptation
デジタルデータあり（Crossref）
すぐに読む
CiNii Research
Differenced maximum mutual information criterion for robust unsupervised acoustic model adaptation
- 資料種別
- 記事
- 著者
- Marc Delcroixほか
- 出版者
- Elsevier BV
- 出版年
- 2016-03
- 資料形態
- デジタル
- 掲載誌名
- Computer Speech & Language 36
- 掲載ページ
- p.24-41
資料詳細
要約等：
- HighlightsThe differenced-MMI (dMMI) is a discriminative criterion that generalizes MPE and BMMI.We discuss the behavior of dMMI when there are errors...
全国の図書館の所蔵
国立国会図書館以外の全国の図書館の所蔵状況を表示します。
所蔵のある図書館から取寄せることが可能かなど、資料の利用方法は、ご自身が利用されるお近くの図書館へご相談ください
書誌情報
この資料の詳細や典拠（同じ主題の資料を指すキーワード、著者名）等を確認できます。
デジタル
- 資料種別
- 記事
- タイトル
- 著者標目
- 出版年月日等
- 2016-03
- 出版年（W3CDTF）
- 2016-03
- タイトル（掲載誌）
- Computer Speech & Language
- 巻号年月日等（掲載誌）
- 36
- 掲載巻
- 36
- 掲載ページ
- 24-41
- 掲載年月日（W3CDTF）
- 2016-03
- ISSN（掲載誌）
- 08852308
- 出版事項（掲載誌）
- Elsevier BV
- 対象利用者
- 一般
- DOI
- 10.1016/j.csl.2015.08.001
- 作成日（W3CDTF）
- 2015-08-19
- 著作権情報
- https://www.elsevier.com/tdm/userlicense/1.0/
- 関連情報（URI）
- 参照
- Discriminative learning for minimum error classification (pattern recognition)Unsupervised Adaptation With Discriminative Mapping TransformsAutomatic determination of acoustic model topology using variational Bayesian estimation and clustering for large vocabulary continuous speech recognitionMaximum likelihood linear regression for speaker adaptation of continuous density hidden Markov modelsStructural Classification Methods Based on Weighted Finite-State Transducers for Automatic Speech RecognitionA Study of Minimum Classification Error (MCE) Linear Regression for Supervised Adaptation of MCE-Trained Continuous-Density Hidden Markov ModelsMPE-based discriminative linear transforms for speaker adaptationDiscriminative Training for Large-Vocabulary Speech Recognition Using Minimum Classification ErrorLightly supervised and unsupervised acoustic model trainingDiscriminative Training for Automatic Speech Recognition: Modeling, Criteria, Optimization, Implementation, and PerformanceCorrectness-Adjusted Unsupervised Discriminative Acoustic Model AdaptationUnsupervised training of acoustic models for large vocabulary continuous speech recognitionAggregate a posteriori linear regression adaptationOn a model-robust training method for speech recognitionMean and variance adaptation within the MLLR frameworkInvestigation of unsupervised adaptation of DNN acoustic models with filter bank inputMaximum mutual information splice transform for seen and unseen conditionsMinimum phone error and I-smoothing for improved discriminative trainingDiscriminative feature transforms using differenced maximum mutual informationN-best based supervised and unsupervised adaptation for native and non-native speakers in carsN-best-based instantaneous speaker adaptation method for speech recognitionAsynchronous stochastic optimization for sequence training of deep neural networks: towards big dataRecognition of conversational telephone speech using the JANUS speech engineDiscriminative speaker adaptation with conditional maximum likelihood linear regressionLattice-based unsupervised MLLR for speaker adaptationModified MMI/MPE: a direct evaluation of the margin in speech recognitionLattice-based optimization of sequence classification criteria for neural-network acoustic modelingMargin-space integration of MPE loss via differencing of MMI functionals for generalized error-weighted discriminative trainingDiscriminative training based on an integrated view of MPE and MMI in margin and error spaceCorpus of spontaneous Japanese: its design and evaluationSpeaker adaptation using lattice-based MLLRDynamic variance adaptation using differenced maximum mutual informationMPE-based discriminative linear transform for speaker adaptationDiscriminative training of language models for speech recognitionSpontaneous speech corpus of JapaneseFlexible discriminative training based on equal error group scores obtained from an error-indexed forward–backward algorithmRecent progress in the MIT spoken lecture processing projectDiscriminative MCE-based speaker adaptation of acoustic models for a spoken lecture processing taskUnsupervised discriminative adaptation using differenced maximum mutual information based linear regressionDiscriminatively trained region dependent feature transforms for speech recognitionA study on soft margin estimation of linear regression parameters for speaker adaptationSoft margin estimation of hidden Markov model parametersInvestigation of full-sequence training of deep belief networks for speech recognitionLarge vocabulary continuous speech recognition based on WFST structured classifiers and deep bottleneck featuresFrame-discriminative and confidence-driven adaptation for LVCSRSequence-discriminative training of deep neural networksImprovements in linear transform based speaker adaptationA unified view for discriminative objective functions based on negative exponential of difference measure between stringsA direct adaptive method for faster backpropagation learning: the RPROP algorithmDiscriminative MAP for acoustic model adaptationDiscriminative linear transforms for speaker adaptationImproved discriminative training using phone latticesBoosted MMI for model and feature-space discriminative trainingIntegrating deep neural networks into structural classification approach based on weighted finite-state transducersUnsupervised training for Mandarin broadcast news and conversation transcriptionfMPE: discriminatively trained features for speech recognition
- 連携機関・データベース
- 国立情報学研究所 : CiNii Research
- 提供元機関・データベース
- Crossref科学研究費助成事業データベース