A fast sequence kernel for sequential data classification (Speech) -- (国際ワークショップ"Asian workshop on speech science and technology")
スポンサーリンク
概要
- 論文の詳細を見る
In this paper, we propose a sequence kernel with fast computation. The kernel is approximately calculated by using a mean vector in feature space. We further studied on log normalization of a sequence kernel to avoid the diagonal dominance problem in this paper. In text-independent speaker identification experiments with 10 male speakers, our approach was not only found to be competitive in identification rates with the conventional sequence kernel, but also achieved two to ten time faster sequence kernel computation.
- 社団法人電子情報通信学会の論文
- 2008-03-13
著者
-
Yamada Makoto
Department Of Chemistry Faculty Of Science Okayama University Of Science
-
Matsui Tomoko
Department Of Statistical Modeling The Institute Of Statistical Mathematics
-
Matsui Tomoko
Department Of Acoustic And Speech Research Advance Telecommunications Research Institute Internation
-
Yamada Makoto
Department Of Chemistry And Biomolecular Science Toho University
-
Yamada Makoto
Department Of Statistical Science The Graduate University For Advanced Studies:center For Advanced S
関連論文
- Improving the Accuracy of Least-Squares Probabilistic Classifiers
- Improving the Accuracy of Least-Squares Probabilistic Classifiers
- The relationships of sleep duration and mental health with electrocardiographic findings : a retrospective-cohort study in Okinawa, Japan
- The relationship between long-term changes in plasma B-type natriuretic peptide levels and electrocardiographic findings
- An Occupational Health Study of Emergency Physicians in Japan : Health Assessment by Immune Variables (CD4, CD8, CD56, and NK Cell Activity) at the Beginning of Work
- Spontaneous Rupture of the Iliac Vein : Report of a Case
- Methods of cross-domain object matching (情報論的学習理論と機械学習)
- Conjugation of Laminin Derived Cell Adhesive Peptides on a Chitosan Membrane and Their Biological Activity
- Screening of Amyloidogenic Peptides in Laminin-1
- Multifunctional Peptide-Chitosan Membrane : New Biomedical Materials
- γ-Functional Prolines Based on Naturally Occurred Hydroxyproline. III
- Accumulation-Exclusion Combined System for the DNA-Binding Harmful Chemicals with Insolubilized DNA
- UV-Irradiated DNA Matrix Selectively Accumulates Heavy Metal Ions
- Effect of Nucleoplasmin on a Nucleosome Structure
- Neurite Outgrowth Promoting Sites on the Laminin Alpha 3 Chain
- Cell Adhesive and Heparin Binding Sites on the Laminin Alpha Chain G Domains
- γ-Functional Prolines Based on Naturally Occurring Hydroxyproline. II
- Identification of Biologically Active Sequences on the Laminin Alpha Chain G Domains
- Identification of Cell Adhesive Sites on the Laminin Alpha Chain Domain VI
- Myonase is Localized in Skeletal Muscle Myofibrils
- Covered Stent Treatment for Traumatic Cervical Carotid Artery Aneurysms : Two Case Reports
- Practical Synthesis of 4-cis Hydroxy-L-Proline
- Ovarian Histological Findings in an Adult Patient with the Steroidogenic Acute Regulatory Protein (StAR) Deficiency Reveal the Impairment of Steroidogenesis by Lipoid Deposition
- Dialogue Speech Recognition by Combining Hierarchical Topic Classification and Language Model Switching(Spoken Language Systems, Corpus-Based Speech Technologies)
- Solid-papillary carcinoma of the breast : Clinicopathological study of 20 cases
- A fast sequence kernel for sequential data classification (Speech) -- (国際ワークショップ"Asian workshop on speech science and technology")
- Information-maximization clustering: analytic solution and model selection (情報論的学習理論と機械学習)
- SERAPH: semi-supervised metric learning paradigm with hyper sparsity (情報論的学習理論と機械学習)
- Direct density-ratio estimation with dimensionality reduction via hetero-distributional subspace analysis (情報論的学習理論と機械学習)
- Estimation of squared-loss mutual information from paired and unpaired samples (情報論的学習理論と機械学習)
- Dependence minimizing regression with model selection for non-linear causal inference under non-Gaussian noise (情報論的学習理論と機械学習)
- Change-Point Detection in Time-Series Data by Relative Density-Ratio Estimation (情報論的学習理論と機械学習)
- Relative Density-Ratio Estimation for Robust Distribution Comparison (情報論的学習理論と機械学習)
- Change-Point Detection in Time-Series Data by Relative Density-Ratio Estimation
- Relative Density-Ratio Estimation for Robust Distribution Comparison
- Prospective Assessment of Pain and Functional Status After Percutaneous Vertebral Body-Perforation Procedure for Treatment of Vertebral Compression Fractures
- Changes in mental health and quality of life with dental implants as evaluated by General Health Questionnaire (GHQ) and Health Utilities Index (HUI)
- Prospective Assessment of Pain and Functional Status After Percutaneous Vertebral Body-Perforation Procedure for Treatment of Vertebral Compression Fractures