SpeechExact (Future)


Speech to Text (STT) can be greatly improved if the STT consults with Cognal’s OntoloNet of Common Terms and semantics.

Manufacturers of tablets and smartphones demand the best Speech-to-Text products. Millions of users obviously depend on the accuracy of the STT on their mobile devices. Typing on a smartphone is cumbersome and to be avoided where possible (except perhaps for gamesters and expert “texters” – e.g., kids).

Manufacturers are attempting to build and sell smartPhones that better understand the user’s meaning of a spoken request.

SpeechExact™ is a perfect fit for Cognal’s Natural Language Understanding (NLU) products.

 SpeechExact: contextual STT 
  • Existing Speech-to-Text (STT) “goes one step further” in choosing the word or phrase
  • Utilizes the same Cognitive Natural Language Understanding algorithms as other products
  • Consults a large and growing taxonomy of ontologies with their terms and semantics (OntoloNet)
  • Should be able to be more accurate in noisy environments (not tested yet)