C
Cheung-Chi Leung
Researcher at Institute for Infocomm Research Singapore
Publications - 64
Citations - 1169
Cheung-Chi Leung is an academic researcher from Institute for Infocomm Research Singapore. The author has contributed to research in topics: Query by Example & Dynamic time warping. The author has an hindex of 22, co-authored 64 publications receiving 1087 citations. Previous affiliations of Cheung-Chi Leung include The Chinese University of Hong Kong & Centre national de la recherche scientifique.
Papers
More filters
Proceedings ArticleDOI
Parallel inference of dirichlet process Gaussian mixture models for unsupervised acoustic modeling: a feasibility study.
TL;DR: Experimental results show that the unsupervised DPGMM posteriorgrams obviously outperformMFCC, and perform comparably to the posterior grams derived from language-mismatched phoneme recognizers in terms of the error rate of ABX discrimination test.
Proceedings ArticleDOI
Joint acoustic modeling of triphones and trigraphemes by multi-task learning deep neural networks for low-resource speech recognition
TL;DR: Estimation of triphone acoustic models in parallel with the estimation of trigrapheme acoustic models under the MTL framework using deep neural network (DNN) shows results that are significantly better than triphone DNNs that are trained by the single-task learning (STL) approach.
Proceedings ArticleDOI
An acoustic segment modeling approach to query-by-example spoken term detection
TL;DR: Experimental results show that the ASMtokenizer outperforms a conventional GMM tokenizer and a language-mismatched phoneme recognizer, and the performance is significantly improved by applying unsupervised speaker normalization techniques.
Journal ArticleDOI
Acoustic segment modeling with spectral clustering methods
TL;DR: This paper uses posterior features as the segment representations, and applies spectral clustering algorithms on the posterior representations of a Gaussian component clustering approach and a segment clustering (SC) approach, which could provide consistent improvement on four different testing scenarios with three evaluation metrics.
Proceedings ArticleDOI
Constrained MLLR for Speaker Recognition
TL;DR: A new feature extraction technique for speaker recognition based on CMLLR speaker adaptation which operates directly on the recorded signal with noise as well as in combination with two cepstral approaches such as reduction in the performance gap between telephone and auxiliary microphone data.