Speech-to-Text

This section describes how to use IDOL Speech Server to perform speech-to-text.

Prepare the Audio

Select the Language Pack

Run the Task

Interpret the Results

CTM Transcript Output

XML Transcript Output

Pauses in Recognition Output

Use a Lattice File

Control Speech-To-Text Process Speed

Enable Word Confidence Scores

Tune Parameters

Noisy Data

Missing Words in Transcript

Extra Words in Transcript

Use a Custom Language Model

Use an Adapted Acoustic Model

Troubleshooting

Generate Diagnostics

Score Speech Recognition

Prepare the Truth Transcript Text

Run the Scorer

Format of Scorer Results

Common Problems Related to Scoring Outputs


_HP_HTML5_bannerTitle.htm