Development

Create Acoustic Model by 'compiling' your speech audio files into a Speaker Dependent Acoustic Model, or adapt the VoxForge Speaker Independent Acoustic Model to better recognize your voice.

Create speaker dependant acoustic model from scratch & run simple dialog manager:
- How-to - using a script
- Tutorial - step-by-step (with explanations)
To improve the recognition rates of the acoustic model you created in the how-to or tutorial, you need to train your acoustic model with many more hours of your speech. A short-cut approach is to 'adapt' the VoxForge speaker independent acoustic model to the characteristics of your voice:
- Adapt Pre-existing Acoustic Model to Your Voice
How to tell how much more speech is needed to improve an acoustic model? You need to test it:
- Testing Your Acoustic Model with HTK & Julius

Create acoustic model & run simple dialog manager:
- How-to - using a script
- Tutorial - step-by-step (with explanations)


Username	Password

Other Dev Pages