Speech Recognition: Theory, Applications, and Challenges


Speech Recognition

Speech Recognition System

Speech Recognition Based on DTW

Hybrid Speech Recognition

Neural Network Speech Recognition

HMM Speech Recognition

Voice Control System

External resources

Advanced Source Code .Com

Neural Networks .It

Face Recognition .It

Iris Recognition .It

Genetic Algorithms .It

Hybrid Speech Recognition System

Download now Matlab source code
Requirements: Matlab, Matlab Signal Processing Toolbox.

The efforts to automate the combination of expert opinions have been studied extensively in the second half of the twentieth century. The combined experts are classifiers and the result of the combination is also a classifier. The outputs of classifiers can be represented as vectors of numbers where the dimension of vectors is equal to the number of classes. As a result, the combination problem can be defined as a problem of finding the combination function accepting N-dimensional score vectors from M classifiers and outputting N final classification scores, where the function is optimal in some sense, e.g. minimizing the misclassification cost.

Combination methods can also be grouped based on the level at which they operate. Combinations of the first type operate at the feature level. The features of each classifier are combined to form a joint feature vector and classification is subsequently performed in the new feature space.

Combinations can also operate at the decision or score level, that is they use outputs of the classifiers for combination. This is a popular approach because the knowledge of the internal structure of classifiers and their feature vectors is not needed. Though there is a possibility that representational information is lost during such combinations, this is usually compensated by the lower complexity of the combination method and superior training of the final system.

We have developed a fast and reliable algorithm for speech recognition for isolated words. The proposed method combines at decision level several algorithms commonly used for speech recognition such as Discrete Cosine Transform, Mel-Frequency Cepstral Coefficients, Linear Predictive Coding, Relative Spectral Transform and Perceptual Linear Prediction. The algorithm for combination can be easily parallelized and run on low-cost hardware in reasonable time.

Index Terms: Matlab, source, code, speech, recognition, isolated, word, words, feature, algorithm, combination, fusion.

Release 1.0 Date 2013.12.16
Major features:
  • Speech recognition based on multiple features selection
  • Optimized algorithm for combination of classifiers
  • Sound acquisition from microphone
  • Sound acquisition from disk
  • Discrete Cosine Transform
  • Mel-Frequency Cepstral Coefficients
  • Linear Predictive Coding
  • Relative Spectral Transform
  • Perceptual Linear Prediction
  • Fast and optimized implementation
  • Easy and intuitive GUI
  • Demo code (protected P-files) available for performance evaluation

Speech Recognition . It Luigi Rosa mobile +39 3207214179 luigi.rosa@tiscali.it