- Subvector-quantized High-density Discrete Hidden Markov Model and its Re-estimation
- PERFORMANCE OF DISCRIMINATIVELY TRAINED AUDITORY FEATURES ON AURORA2 AND AURORA3
- Fast GMM Computation for Speaker Verification Using Scalar Quantization and Discrete Densities
- PHONE DELETION MODELING IN SPEECH RECOGNITION
- FAST SPEAKER ADAPTION VIA MAXIMUM PENALIZED LIKELIHOOD KERNEL REGRESSION
- Robust Speaker Verification Using Short-Time Frequency with Long-Time Window and Fusion of Multi-Resolutions
- DISCRIMINATIVE TRAINING OF STREAM WEIGHTS IN A MULTI-STREAM
- KERNEL EIGENSPACE-BASED MLLR HSIAO WEND HUU, ROGER
- -15 -10 -5 0 5 10 15 Dimension 1
- IEEE TRANSACTIONS ON AUDIO, SPEECH AND LANGUAGE PROCESSING, March 17, 2009 1 Maximum Penalized Likelihood Kernel Regression
- IEEE SIGNAL PROCESSING LETTERS 1 Minimization of Utterance Verification Error Rate
- Joint Optimization of the Frequency-domain and Time-domain Transformations in Deriving
- PASSENGER ROUTE GUIDANCE SYSTEM FOR MULTI-MODAL TRANSIT NETWORKS
- IEEE TRANSACTIONS ON SPEECH AND AUDIO PROCESSING, September 7, 2004 1 Kernel Eigenvoice Speaker Adaptation
- An Acoustic-Phonetic and a Model-Theoretic Analysis of Subspace Distribution Clustering Hidden Markov Models
- IEEE TRANSACTIONS ON SPEECH AND AUDIO PROCESSING. 1 Subspace Distribution Clustering Hidden Markov
- EIGENTRIPHONES: A BASIS FOR CONTEXT-DEPENDENT ACOUSTIC MODELING Tom Ko and Brian Mak
- The Use of Subvector Quantization and Discrete Densities for Fast GMM Computation for Speaker Verification
- Automatic Estimation of Decoding Parameters Using Large-Margin Iterative Linear Programming
- Min-max Discriminative Training of Decoding Parameters Using Iterative Linear Programming
- Unsupervised Speaker Adaptation using Reference Speaker Weighting
- KERNEL EIGENSPACE-BASED MLLR ADAPTATION USING MULTIPLE REGRESSION Roger Hsiao and Brian Mak
- VARIOUS REFERENCE SPEAKERS DETERMINATION METHODS FOR EMBEDDED KERNEL EIGENVOICE SPEAKER ADAPTATION
- DISCRIMINATIVETRAINING OF AUDITORY FILTERS OF DIFFERENT SHAPES FOR ROBUST SPEECH RECOGNITION
- KNOWLEDGE-BASED SENSE PRUNING USING THE HOWNET: AN ALTERNATIVE TO WORD SENSE DISAMBIGUATION
- DISCRIMINATIVE AUDITORY FEATURES FOR ROBUST SPEECH RECOGNITION Brian Mak, Yik-Cheung Tam
- AN ALTERNATIVE APPROACH OF FINDING COMPETING HYPOTHESES FOR BETTER MINIMUM CLASSIFICATION ERROR TRAINING
- Development of an Asynchronous Multi-band System for Continuous Speech Recognition
- Rapid Speaker Adaptation Using MLLR and Subspace Regression Classes Kwok-Man Wong, Brian Mak
- Eigenvoice Speaker Adaptation via Composite James T. Kwok, Brian Mak and Simon Ho
- DISCRIMINATIVE TRAINING BY ITERATIVE LINEAR PROGRAMMING OPTIMIZATION Brian Mak, Benny Ng
- Improving Eigenspace-based MLLR Adaptation by Kernel PCA Brian Mak and Roger Hsiao
- A STUDY OF VARIOUS COMPOSITE KERNELS FOR KERNEL EIGENVOICE SPEAKER Brian Mak, James T. Kwok, and Simon Ho
- Pruning Hidden Markov Models with Optimal Brain Surgeon
- Knowledge-based Sense Pruning using the an Alternative to Word Sense Disambiguation
- Problems of Modeling Phone Deletion in Conversational Speech for Speech Recognition
- IEEE TRANSACTIONS ON SPEECH AND AUDIO PROCESSING, June 3, 2006 1 Kernel Eigenspace-based MLLR Adaptation
- IEEE TRANSACTIONS ON SPEECH AND AUDIO PROCESSING, September 20, 2003 1 Discriminative Auditory-based Features for
- IMPROVING REFERENCE SPEAKER WEIGHTING ADAPTATION BY THE USE OF MAXIMUM-LIKELIHOOD REFERENCE SPEAKERS
- DISCRIMINATIVE FEATURE TRANSFORMATION BY GUIDED DISCRIMINATIVE Roger Hsiao and Brian Mak
- IEEE TRANSACTIONS ON SPEECH AND AUDIO PROCESSING, December 21, 2005 1 Embedded Kernel Eigenvoice Speaker Adaptation
- A COMPARISON OF VARIOUS ADAPTATION METHODS FOR SPEAKER VERIFICATION WITH LIMITED ENROLLMENT DATA
- subspace distribution clustering HMM
- IMPROVING SPEECH RECOGNITION BY EXPLICIT MODELING OF PHONE DELETIONS Tom Ko, Brian Mak
- A FULLY AUTOMATED DERIVATION OF STATE-BASED EIGENTRIPHONES FOR TRIPHONE MODELING WITH NO TIED STATES USING REGULARIZATION
