Voice Text Input Widget

Voice Text Input technology allows free-form speech to be transcribed into text, supporting short or long utterances, with or without numbers.

Widget Navigation

Let’s start by exploring the widget's main options.

Navigate back (1) to the Project Hub
Change the Selected Model (2) you are editing
Widget global editing tools (3):
- Add this model to Favorites
- Create a new model
- Rename the model
- Configure the recognizer
- Delete the model
Change the Model’s Language (4)
Choose an optional Speech Enhancement model (5) to improve voice detection in noisy environments
Quick Test (6) your model in real time by speaking directly in the widget

Testing the Voice Text Input Model

When the test panel is open and ready, click Start recording (1) to play the model for one minute. Adjust the confidence threshold (2) in real time during the test to display only hypotheses that meet the minimum confidence score. Speak your voice commands aloud to see the hypotheses appear (3) according to your model's recognizer parameters.

What are the recognizer options ?

Based on this documentation page containing Recognizer parameters, we provide the most used parameters that have a real impact in Quick Testing. Remember you can still change those parameters after Downloading the project.

TSILENCE: The silence period needed to detect the spoken sentence has ended. Any voice recognized afterward will be treated as a new sentence.

STREAM_RESULT_MODE: This means the text appears as you speak. Partial sentences will appear even if you do not pass the TSILENCE.

MAXNBEST: This value specifies the maximum number of hypotheses shown in the test panel. Transcript history contains only the main hypothesis.