Up to 99% recognition accuracy with only minimal training required!
This myth has been perpetuated by the developer himself in order to sell more software. Recognition improves with usage. Realistically and of particular importance in a professional environment, the software requires the following pre-adaptation in order to minimize corrections by its user:
- Creation of a user’s particular acoustic profile – how he or she pronounces all the phonemes in the desired language.
- Creation of an appropriate dictionary which contains all the vocabulary which will be used – if the word or name is not in the active dictionary, it cannot be recognized.
- Incorporation of the individuals writing style from previously created documents – context adaptation provides additional information to the software and significantly improves the recognition rate.
Speech recognition is intuitive and simple to learn and use!
Without a structured, comprehensive and customized training, speech recognition can be tedious to learn and in many cases the initial investment is lost as the software is abandoned by the user.
Specialists, knowledgeable in every aspect of the software provide a solid foundation upon which the user(s) can build on.
I tried speech recognition before. It didn’t work then and it likely won’t work now!
While this may have been true several years ago, great strides have been made in the following three areas:
- The software has become much more user-friendly with more intuitive commands coupled with a narrow integration with popular applications such as Microsoft Word and Outlook as well as Explorer.
- Computer processors are much more powerful today and provide the speech recognition engine with sufficient resources required to analyse and recognize speech effectively.
- As speech recognition is being deployed by more and more professionals, integrators with a deep knowledge of the software and the expertise to integrate the technology into various different workflows and applications have greatly improved the user experience.