Connectionist speaker normalization and its applications to speech recognition
Published 9 December 2002
Xuedong Huang, K.F. Lee, Alexander Waibel
Citations8
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
A codeword-dependent neural network is presented as a nonlinear mapping function to transform speech data between two speakers that significantly reduced the error rate and made full use of dynamic information.
Abstract
Speaker normalization may have a significant impact on both speaker-adaptive and speaker-independent speech recognition. In this paper, a codeword-dependent neural network (CDNN) is presented for speaker normalization. The network is used as a nonlinear mapping function to transform speech data between two speakers. The mapping function is characterized by two important properties. First, the assembly of mapping functions enhances overall mapping quality. Second, multiple input vectors are used simultaneously in the transformation. This not only makes full use of dynamic information but also alleviates possible errors in the supervision data. Large-vocabulary continuous speech recognition is chosen to study the effect of speaker normalization. Using speaker-dependent semi-continuous hidden Markov models, performance evaluation over 360 testing sentences from new speakers showed that speaker normalization significantly reduced the error rate from 41.9% to 5.0% when only 40 speaker-dependent sentences were used to estimate CDNN parameters.>
Keywords
Computer Science
Learning Internal Representations by Error Propagation
16,235 Citations1985David E. Rumelhart, Geoffrey E. Hinton +1 more
IEEE Transactions on CommunicationsAn Algorithm for Vector Quantizer Design
7,183 Citations1980Y. Linde, A. Buzo +1 more
An efficient and intuitive algorithm is presented for the design of vector quantizers based either on a known probabilistic model or on a long training sequence of data.
IEEE Transactions on Acoustics Speech and Signal ProcessingDynamic programming algorithm optimization for spoken word recognition
6,544 Citations1978Hiroaki Sakoe, Seibi Chiba
This paper reports on an optimum dynamic progxamming (DP) based time-normalization algorithm for spoken word recognition, in which the warping function slope is restricted so as to improve discrimination between words in different categories.
IEEE Transactions on Acoustics Speech and Signal ProcessingAn overview of the SPHINX speech recognition system
439 Citations1990K.-F. Lee, Hsiao-Wuen Hon +1 more
SPHINX is a system that demonstrates the feasibility of accurate, large-vocabulary, speaker-independent, continuous speech recognition, based on discrete hidden Markov models with LPC- (linear-predictive-coding) derived parameters.
Voice conversion through vector quantization
417 Citations2003Masanobu Abe, Satoshi Nakamura +2 more
The authors propose a new voice conversion technique through vector quantization and spectrum mapping which makes it possible to precisely control voice individuality.
International Conference on Acoustics, Speech, and Signal ProcessingEnvironmental robustness in automatic speech recognition
275 Citations2002Alejandro Acero, Richard M. Stern
Initial efforts to make Sphinx, a continuous-speech speaker-independent recognition system, robust to changes in the environment are reported, and two novel methods based on additive corrections in the cepstral domain are proposed.
International Conference on Acoustics, Speech, and Signal ProcessingConsonant recognition by modular construction of large phonemic time-delay neural networks
153 Citations2003Alexander Waibel, Hirozumi Sawai +1 more
It is shown that neural networks for speech recognition can be constructed in a modular fashion by exploiting the hidden structure of previously trained phonetic subcategory networks to avoid the excessive learning times that would be necessary to train larger networks and allow for incremental learning.
Journal of the Acoustical Society of Japan (E)Voice conversion through vector quantization.
151 Citations1990Masanobu Abe, Satoshi Nakamura +2 more
Noise reduction using connectionist models
132 Citations2003Shinichi Tamura, Alex Waibel
Using a back propagation network learning algorithm, a four-layered feed-forward network is trained on learning samples to realize a mapping from the set of noisy signals a set of noise-free signals, showing that the network can indeed learn to perform noise reduction.
International Conference on Acoustics, Speech, and Signal ProcessingSpeaker-independent word recognition using a neural prediction model
85 Citations2002K.-I. Iso, Takao Watanabe
A speech recognition model called the neural prediction model (NPM) is proposed, which uses a sequence of multilayer perceptrons (MLPs) as a separate nonlinear predictor for each class to represent temporal structures of speech patterns as recognition cues.
Improved acoustic modeling with the SPHINX speech recognition system
48 Citations1991Xuedong Huang, K.F. Lee +2 more
Several techniques that substantially reduced SPHINX's error rate are presented, which include dynamic features, semicontinuous hidden Markov models, speaker clustering, and the shared distribution modeling.
DARPA resource management benchmark test results June 1990
30 Citations1990David S. Pallett, J. Fiscus +1 more
The June 1990 DARPA Resource Management Benchmark Test makes use of the first of several test sets provided with the Extended Resource Management Speaker-Dependent Corpus, designed as a speaker-dependent extension to the Resource Management (RM1) Corpus.
International Conference on Acoustics, Speech, and Signal ProcessingFast speaker adaptation for speech recognition systems
29 Citations2002Fritz Class, A. Kaltenmeier +2 more
Different speaker adaptation methods for speech recognition systems adapting automatically to new and unknown speakers in a short training phase are discussed, and the results show that in both systems speaker-adaptive error rates are close to speaker-dependent error rates.
Acoustic to articulatory parameter mapping using an assembly of neural networks
29 Citations1991Mazin G. Rahim, W.B. Keijn +2 more
An assembly of multilayer perceptrons, each designated to a specific region in the articulatory space, is used to map acoustic parameters of the speech into tract areas using neural networks.
International Conference on Acoustics, Speech, and Signal ProcessingA comparative study of spectral mapping for speaker adaptation
24 Citations2002Satoshi Nakamura, Kiyohiro Shikano
An investigation of the actual spectrogram shows that the fuzzy VQ-based spectral mapping works better than the neural network mapping, which is about 6% better in spectral distortion than nonlinear spectral mapping by a feedforward neural network.
Neutral nets for computing
23 Citations2003Richard P. Lippmann
There has been a resurgence of interest in neutral net models composed of many simple interconnected processing elements operating in parallel, and a major emphasis is placed on relating these models to existing classification and clustering algorithms.
Spectral transformations through canonical correlation analysis for speaker adptation in ASR
22 Citations2005Khalid Choukri, Gérard Chollet +1 more
This paper describes a technique of spectral transformation for improved adaptation of a knowledge data base or reference templates to new speakers in automatic speech recognition (ASR).
International Conference on Acoustics, Speech, and Signal ProcessingThe Meta-Pi network: connectionist rapid adaptation for high-performance multi-speaker phoneme recognition
20 Citations2002J.B. Hampshire, Alex Waibel
The Mega-Pi paradigm implements a dynamically adaptive Bayesian MAP classifier and the Meta-Pi model is a viable basis for a connectionist speech recognition system that can rapidly adapt to new speakers and varying speaker dialects.
A study on speaker-adaptive speech recognition
19 Citations1991Xuedong Huang
In this paper, DARPA Resource Management task is used as the domain to investigate the performance of speaker-adaptive speech recognition, and preliminary results indicate that speaker differences can be well minimized.
On speaker-independent, speaker-dependent, and speaker-adaptive speech recognition
17 Citations1991Xuedong Huang, K.F. Lee
Application of neural networks to articulatory motion estimation
16 Citations1991Tetsuo Kobayashi, Mitsuhiko Yagyu +1 more
An application of neural networks (NNs) to the problem of estimating the motion of articulatory organs from speech waves was successfully applied and a high-speed and stable articulatory parameter estimation technique can be realized using neural networks.
International Conference on Acoustics, Speech, and Signal ProcessingSpeech recognition using temporal decomposition and multi-layer feed-forward automata
12 Citations2003Claude Montacié, Khalid Choukri +1 more
Two series of experiments are reported on using multilayer feed-forward automata (MLFFA) to control some aspects of intraspeaker and interspeaker variability in automatic speech recognition.