What is TalkingPoint SpeechToText?
Posted: 16/03/2007
Question: What is TalkingPoint SpeechToText?
Answer: SpeechToText is a speech recognition product designed to streamline the use of Digital Dictation in a professional organisation. Sound recordings from TalkingPoint SoundRecorder or a Digital Recorder are saved to a network for transcription by secretarial staff. Automatic SpeechToText provides batch SpeechRecognition. This assists the secretary by converting the sound file to text, leaving only correction and finishing off to be done.
Once the text is corrected there are facilities for automatically placing the finished text into a word document.
The system also supports pre-defined variables. This allows the author to specify the name and value of the variable by voice (e.g. Client Name). These variables are then automatically entered into named fields in the word document. This can streamline workflows such as Clinic Reports or Building Surveys. The author specifies the variables and they are automatically filled in in the resulting form.
The process works in batch, automatically loading the correct voice files for each author's dictation. Correction by the secretary is also used to improve the author's voice files so that recognition accuracy improves over time. Facilities are also provided for batch processing of previous reports to improve recognition by updating the context statistics for the author's voice files.
