Speak&Find Features

Monitoring: the user can check new recordings for specific keywords. This type of monitoring means that the user can be alerted when a particular word is found in newly added sound data.

Transcription: the user can see the spoken text of the added sound data. Similarly you can also see texts of already stored sounds.

Phoneme recognition: the user can define the pronunciation of keywords or phrases simply by speaking. Speak&Find translates the spoken sound into the appropriate phonemes and uses them for future searches or indexes.

Specialized Vocabularies: Speak&Find supports Philips ConTexts. ConTexts are specialized vocabularies which guarantee the best possible accuracy for your domain. You can use any Philips certified ConText

Speaker Training: if the speakers of the recorded audio data are known the technology can be trained to the speaker's voice. The Speak&Find API contains a complete GUI (Graphical User Interface) to facilitate the integration of user training. 

Caching: additionally the positions of words or phrases are cached. This means that the system only has to search for that word or phrase once. 

Database independency: the Speak&Find Technology system is independent of the database you are using and can be easily tailored to any current or future database solution.

Speaker independency: the Speak&Find Technology can handle an unlimited number of different speakers. However the searching accuracy and speed can be greatly improved when only one speaker's data is stored in the database and speaker training can be applied.

Network distribution: the Speak&Find Technology core system runs on a PC based Windows system. However, it is designed so that it can work across any network.