Jan 06

Interactive dialogue

Over the past year , starting with Google glasses , the rise of wearable devices , intelligent home and automotive equipment , voice recognition technology to the foreground application .
Speech recognition for the entire industry, the past year has experienced unprecedented prosperity , speech recognition technology has greatly improved, especially for emerging entrepreneurial companies, voice-recognition technology is no longer as closed as in the past , but the gradual opening and open source , voice technology threshold decreased. “Professional ‘s voice recognition can do 90% of small firms can take advantage of open source technology can do at least about 70 % recognition rate .” Speech Recognition insiders such judgments .
In fact, even as a professional speech recognition vendors IFLYTEK , cloud -known sound, so you can think Chi will be more than 90% speech recognition accuracy , voice recognition alone in the application of wearable devices are still not ” obedient” . It even became one of the biggest pain points overall speech technology applications.
“Speech Recognition regret is again trying to do can not do one hundred percent .” Intelligent speech technology start-up companies will think the chief scientist Yu Kai Chi such judgments .
The Royal Academy of Engineering , University of Cambridge Professor Steve Young is the authority figure of speech recognition industry , many years ago to start thinking in another direction : “In recognition error or ambiguity in the understanding of the situation , you can continue dialogue and communication between people achieve their goals , the machine should be done . ”
Along the ” dialogue” technology path laid in the UK , as Steve Early dialogue system core research team member Yang , speech recognition , University of Cambridge , Dr. Yu Kai, and another in Cambridge graduate with a high shixing home business , founded the Chi will think . Chi will think from the perspective of progress , not perfection entangled in recognition rate of speech recognition , and human-computer interaction Sike “dialogue” system seems to be how to become a truly wearable devices can be used on voice technology “shortcut” one .