February 15, 2004 - Changed from LGPL to BSD license and various updates.

This is the first update in a while. OCV license changed from LGPL to BSD license, which gives much more freedom. The links page had several dead links, and they are either fixed or removed. I (Danny) am starting to work on improvements for original OCV code, but this time coding it in C. There may be a Java API once and if I do finish. My aim is to have a working Hidden Markov Model module for training and try to make the recognizer less speaker dependent.

June 28, 2002 - First release of complete package

We release the complete package to the public. Although this initial package contains only an Isolated Word Recognizer, it will become more sophisticated as we add more capability to the engine.

May 23, 2002 - Completion of Vector Quantization component

We finished programming the Feature Extraction component and use it to extract Mel-Frequency Cepstral Coefficients from speech signal.

May 16, 2002 - Completion of Feature Extraction component

We finished programming the Feature Extraction component and use it to extract Mel-Frequency Cepstral Coefficients from speech signal.

May 5, 2002 - Completion of Hidden Markov Models component

We finished programming the Discrete Hidden Markov Models for our speech recognition engine. EDIT: It was found that our HMM module can't be used for recognition. We need better understanding of HMM.

March 28, 2002 - Project started

We started doing preliminary research on speech recognition.