Click here to register.

Adapt Speaker Independent Acoustic Model to Your Voice

In the VoxForge Acoustic Model Creation How-to and/or Tutorial, you created a Speaker Dependant Acoustic Model with a small amount of transcribed speech (using your voice). 

To improve your Acoustic Model, you need to record many additional hours of your speech and 're-compile' your Acoustic Models to improve your recognition accuracy.  An alternative to this is to 'adapt' the VoxForge Speaker Independent Acoustic Model to the characteristics of your voice.

This Tutorial walks you through the steps required to take the VoxForge Speaker Independent Acoustic Model and adapt it to your voice using the Acoustic Models you created in the How-to or Tutorial.