An LVQ based reference model for speaker-adaptive speech recognition
- 1 January 1992
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 1 (15206149) , 441-444 vol.1
- https://doi.org/10.1109/icassp.1992.225877
Abstract
A novel type of hierarchical phoneme model for speaker adaptation, based on both hidden Markov models (HMM) and learned vector quantization (LVQ) networks is presented. Low-level tied LVQ phoneme models are trained speaker-dependently and independently, yielding a pool of speaker-biased phoneme models which can be mixed into high-level speaker-adaptive phoneme models. Rapid speaker adaptation is performed by finding an optimal mixture for these models at recognition time, given only a small amount of speech data; subsequently, the models are fine-tuned to the new speaker's voice by further parameter reestimation. In preliminary experiments with a continuous speech task using 40 context-free phoneme models at task perplexity 111, the authors achieved 82% word accuracy for speaker-dependent recognition and 73% in the speaker-adaptive mode.Keywords
This publication has 6 references indexed in Scilit:
- The Meta-Pi network: connectionist rapid adaptation for high-performance multi-speaker phoneme recognitionPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- A new paradigm for speaker-independent trainingPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1991
- Speaker-independent large vocabulary word recognition using an LVQ/HMM hybrid algorithmPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1991
- Continuous speech recognition using linked predictive neural networksPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1991
- Statistical pattern recognition with neural networks: benchmarking studiesPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1988
- Structural methods in automatic speech recognitionProceedings of the IEEE, 1985