Vector Quantizer Signal Transform
نویسندگان
چکیده
This paper deals with the problem of combination of Neural Networks (NN) and traditional statistical pattern classiiers. It is shown that a Neural Network can be used to replace the vector quantizer (VQ) and some feature extraction and feature reduction modules in a discrete pattern recognition system. A criterion for training the NN-weights and the classiier jointly is derived, leading to the maximum mutual information (MMI)-paradigm for NN-training. NN-parameter optimization can be performed by the MMI-Net that is trained by a new proposed gradient based learning approach. As an application a quite simple MMI-Net is integrated in a large vocabulary speech recognition system. This new hybrid speech recognition system achieved State-of-the-Art performance for the 1000 word speaker independent ARPA Resource Management database resulting in an average word recognition rate of 95,2%.
منابع مشابه
The Coding-Optimal Transform
We propose a new transform coding algorithm that integrates all optimiza tion steps into a coherent and consistent framework Each iteration of the algorithm is designed to minimize coding distortion as a function of both the transform and quantizer designs Our algorithm is a constrained version of the LBG algorithm for vector quantizer design The reproduction vectors are constrained to lie at t...
متن کاملDispersion phase vector quantization for enhancement of waveform interpolative coder
This paper presents an efficient analysis-by-synthesis vector quantizer for the dispersion phase of the excitation signal which was used to enhance a waveforminterpolative coder. The scheme can be used to enhance other harmonic coders, such as the sinusoidal-transform coder and the multiband-excitation coder. The scheme incorporates perceptual weighting, and does not require any phase unwarping...
متن کاملEnergy-Based Split Vector Quantizer Employing Signal Representation in Multiple Transform Domains
متن کامل
Two stage transform vector quantization of LSFs for wideband speech coding
We investigate the use of a two stage transform vector quantizer (TSTVQ) for coding of line spectral frequency (LSF) parameters in wideband speech coding. The first stage quantizer of TSTVQ, provides better matching of source distribution and the second stage quantizer provides additional coding gain through using an individual cluster specific decorrelating transform and variance normalization...
متن کاملHigh-resolution quantization theory and the vector quantizer advantage
How much performance advantage can a fixed dimension vector quantizer gain over a scalar quantizer? We collect several results from high-resolution or asymptotic (in rate) quantization theory and use them to identify source and system characteristics that contribute to the vector quantizer advantage. One well-known advantage is due to improvement in the space-filling properties of polytopes as ...
متن کاملAn algorithm for uniform vector quantizer design
A vector quantizer maps a k-dimensional vector into one of a finite set of output vectors or “points”. Although certain lattices have been shown to have desirable properties for vector quantization applications, there are as yet no algorithms available in the quantization literature for building quantizers based on these lattices. An algorithm for designing vector quantizers based on the root l...
متن کامل