Reiser - Dedikert sensor for talegjenkjenning

Speech recognition technology is reaching product maturity and is becoming a mainstream, preferred mode of interaction, both in consumer and enterprise markets. Microsoft Xbox, Google Now, iPhone Siri and Amazon Echo are some examples showcasing the trend. BCC Research predicts the global speech recognition market will reach $113 billion in 2017, after increasing at a five-year compound annual growth rate of 16.2%. Over the last decade, a significant source of improvement in voice recognition rates has come from better and more powerful software. Algorithms have been improved, and access to the compute power needed to crunch the signals and detect and classify the spoken words correctly is enabled via cloud computing. More recently however, established companies as well as startups have turned their attention to hardware components and that can boost the performance of speech recognition, improving on speed, accuracy and robustness. In this project, we develop and evaluate a new hardware component from SINTEF MiNaLab with a significant and potentially game-changing impact on speech recognition technologies in the years to come.

Det søkes om reisestøtte til 4 reiser til USA. Formålet er å videreføre arbeidet med å kommersialisere SINTEF's optiske mikrofon. Gjennom prosjektet Array-on-a-Dime har vi utforsket markedet for optiske MEMS-mikrofoner for bruk til bedre talegjenkjenning, og fått stor interesse hos store amerikanske bedrifter.