Self-learning vector quantization for pattern discovery from speech
نویسندگان
چکیده
A novel and computationally straightforward clustering algorithm was developed for vector quantization (VQ) of speech signals for a task of unsupervised pattern discovery (PD) from speech. The algorithm works in purely incremental mode, is computationally extremely feasible, and achieves comparable classification quality with the well-known k-means algorithm in the PD task. In addition to presenting the algorithm, general findings regarding the relationship between the amounts of training material, convergence of the clustering algorithm, and the ultimate quality of VQ codebooks are discussed.
منابع مشابه
The Time Adaptive Self Organizing Map for Distribution Estimation
The feature map represented by the set of weight vectors of the basic SOM (Self-Organizing Map) provides a good approximation to the input space from which the sample vectors come. But the timedecreasing learning rate and neighborhood function of the basic SOM algorithm reduce its capability to adapt weights for a varied environment. In dealing with non-stationary input distributions and changi...
متن کاملUsing Self-organizing Maps and Learning Vector Quantization for Mix- Ture Density Hidden Markov Models Using Self-organizing Maps and Learning Vector Quanti- Zation for Mixture Density Hidden Markov Models. Acta Polytechnica
Thesis for the degree of Doctor of Technology to be presented with due permission for public examination and criticism in Auditorium F1 of the Helsinki University of Technology on the 3rd of October, at 12 o'clock noon. ABSTRACT This work presents experiments to recognize pattern sequences using hidden Markov models (HMMs). The pattern sequences in the experiments are computed from speech signa...
متن کاملInfluence of Vector Quantization on Isolated Word Recognition
Vector Quantization can be considered as a data compression technique. In the last few years, vector quantization has been increasingly applied to reduce problem complexity like pattern recognition. In speech recognition, discrete systems are developed to build up real-time systems. This paper presents original results by comparing the KMeans and the Kohonen approaches on the same recognition p...
متن کاملUsing FFI Interpolator and VQ Quantization for Designing of High Quality 1200 BPS Speech Vocoder
Storaging or transmission of speech signals at very low bit rate is a hot area in the field of speech processing. We used stochastic inter-frame interpolators and vector quantization (VQ) as a new method for developing a high quality 1200 BPS speech vocoder. The objective and subjecgtive test results show that performance of the new vocoder is compairable with 4800 BPS standard vocoders (as CELP).
متن کاملSelf-organization in mixture densities of HMM based speech recognition
In this paper experiments are presented to apply Self-Organizing Map (SOM) and Learning Vector Quantization (LVQ) for training mixture density hidden Markov models (HMMs) in automatic speech recognition. The decoding of spoken words into text is made using speaker dependent, but vocabulary and context independent phoneme HMMs. Each HMM has a set of states and the output density of each state is...
متن کامل