Workflow of OpenDL Speech Recognition Solution

With SpeechMagic, your IT provider can build speech recognition applications to suit your individual needs. Some institutions prefer to use their existing workflows which have proven optimal for their situation. Others are interested in working with new, highly efficient scenarios. Whatever your goals, SpeechMagic will help you get there.

Click the picture to see various SpeechMagic scenarios, or watch our short film explaining the workflow (Flash file).

Dictation

The first step in the document creation process is to record a dictation, which can be done with a variety of recording devices. SpeechMagic can process sound files recorded with PC-compatible headsets and microphones such as Philips SpeechMike as well as mobile input devices such as digital recorders, e.g. Philips DPMs, PDAs, Tablet PC, and standard landline telephones. A sound file can also be recorded and sent over the Internet or a WAN to SpeechMagic.

Recognition

SpeechMagic automatically converts the audio data into a full document in the format you are used to. The document can be accessed from any PC.

Editing

Secretaries review the recognized text and correct recognition errors. Alternatively, you can correct reports yourself or outsource to an external transcription service.

Learning

The system analyzes the corrected texts automatically, becomes familiar with how you speak and your language model is automatically adapted (without your further interaction). New words are easily added to the ConText.