1 / 14

Speech Recognition through Neural Networks

Speech Recognition through Neural Networks. By Mohammad Usman Afzal Mohammad Waseem. Introduction.

lynnea
Download Presentation

Speech Recognition through Neural Networks

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Speech Recognition through Neural Networks By Mohammad Usman Afzal Mohammad Waseem

  2. Introduction • Speech Recognition is a process by which a computer maps an acoustic speech signal to text. Speech recognition has several stages. Its stages involve digital sampling of speech, acoustic signal processing and generating coefficients. The final stage is the recognition of phonemes, groups of phonemes and words. Multi Layered Feed Forward Perceptron Neural Network is used to generate output.

  3. Speech Production

  4. Speech Recognition Techniques • Feature Extraction • Artificial Intelligence • Pattern Recognition

  5. System Overview

  6. Speech Recognizer

  7. Digitization of speech • Recording the sound • Analog to Digital converter • Sampling and Quantization

  8. Filtering the Signal • Aliasing • Eliminate the signal over half the sampling frequency (that is SF/2) • Eliminating Pure Noise and Silence • Pure Noise and Silence is eliminated using Zero Crossing Rate (ZCR) and Power of the signal

  9. Classification of Signal

  10. Coefficients Measurement • Need of Coefficients • To get rid of a lot of digital signal data • Advantage of LP Coefficients • Applicable to time domain as well as frequency domain

  11. LP Coefficients

  12. Feed Forward Perceptron Neural Networks • Input Layer • 8 LP Coefficients fed to the input layer • Hidden Layer • Output Layer • Generates the output according to the weights assigned

  13. Neural Network

  14. Future Work • Complete Urdu Speech Recognizer • Interface with an Urdu Editor • Interface with Urdu to English Translator

More Related