Application
|
Dictation rules
|
Human language features a very complex structure, being humanised by each individual, influencing it by the pitch of his/her voice, local dialect, gender, clear pronunciation and speech rate, but also by moods and situations.
Although modern speech recognition systems already perform amazing results, you should keep in mind that they don’t offer plug&play solutions. Therefore it is advisable to attend instruction courses before using the device, in any case.
The recognition rate of the speech recognition system can be influenced by many different reasons (background noise, dialects, fast or unclear pronunciation, loss due to .dss compression, the input microphone, etc.) Therefore some important rules should be taken into account to achieve optimum speech recognition support:
-
Start your own voice profile and train it continuously. Thus the
program learns gradually to convert best your voice and your specific
vocabulary into typed text.
-
A few speech recognition programs already offer specific vocabulary,
for example for medical staff or lawyers, thus shortening the learning
phase and facilitating to start the new application.
-
Please check the regulation of your device’s microphone sensitivity as well as the recording’s volume control in Windows.
-
Please pay attention to the microphone distance. The voice distance to the microphone should be 5-10 cm.
-
If possible do not speak directly into the microphone, to avoid annoying sibilants drowning out the text.
-
Please speak as uniformly and rhythmic as possible (voice level and accent).
-
Pay attention that your working environment is as calm as possible (avoid background noise).
|