To achieve success with speech recognition, you must train the speech recognition tool to recognize your voice and train yourself to speak clearly and consistently into the microphone. Unlike a person, who can interpret a word or phrase both by its sound and by its context, speech recognition is literal and requires a clear articulation of every phrase. Consistent, clear speech is essential to achieve any sort of accuracy. Assuming you just installed speech recognition, the Voice Training Default Speech Profile dialog box should appear on your screen, as shown in Figure 15.6, indicating that you must train speech recognition before you begin using it. If this dialog box does not appear, or if you ran a training session and want to run it again, click the Tools button in the Language bar and choose Training . The Voice Training Default Speech Profile dialog box appears, welcoming you to the training session. Click Next and follow the onscreen instructions to position your microphone properly and test it. Figure 15.6. You must complete one training session before using speech recognition.
After you complete the microphone setup, the Voice Training Wizard should appear. If the Voice Training Wizard is not onscreen, or if you already ran it and want to run it again (to fine-tune its operation), you can run it from the Language bar. Click the Tools button in the Language bar and click Training . Follow the wizard's instructions to complete the training session. The wizard displays a series of dialog boxes that require you to read sentences into your microphone, as shown in Figure 15.7. Figure 15.7. The Voice Training Wizard prompts you to read aloud .
Caution
The entire session takes about 15 minutes. Here are some tips to make the training session more successful:
If you share a computer with other users, a computer trained for your voice obviously will be less responsive to other voices. Fortunately, each user can train speech recognition for his or her own voice by creating a separate profile. To create a profile, perform the following steps:
After you have set up two or more recognition profiles, you easily can switch from one profile to another. Simply click the Language bar's Tools button, point to Current User , and click the name of the desired user (or click Default Speech Profile to use the profile you created when you ran and trained speech recognition for the first time). Note
|