Speech Recogition 2008-2009

Evolutionary Computing & AI Project,

This is my first project. I begin to deal with it when I was freshman. I have learned program in High School. So I choosen Speech Recognition as my first project. I want to learn more things. The main work of this project is try to learn methods about speech recognition, patteren recognition, machine learning and some basic method to deal with very complex optimization.

Main Objection

Collecting all kinds of algorithm about how to design speech recognition software and creating software for speech recognition.

The methods I applied

1. Distance Method: Dyamic Time Wrap

2. feature extraction: MFCC, LPCC, Simple Wavelet Decompression ICA

3. Reducing dimension: PCA

4. Reducing noise: median filter, lowpass filter, FIR

5. Main Model: Hidden Markov Model

6. Cluster Model: K-means, FCM, VQ

7.Trying Model: BP-ANN, Simple GA

8.Preprocessing: VAD method

Partly Work:

The feature I used in early stage in this project:

Interface:

Download Reference

Return to Research Page