MIT develops specialised low power speech recognition chip for electronics 2017-02-22 17:15:39 / CONFERENCES

MIT announced today that it’s developed a speech recognition chip capable of real world power savings of between 90 and 99 percent over existing technologies. Voice technology has, of course, become nearly ubiquitous in mobile devices, thanks to the exponential growth of smart assistants like Siri, Alexa and Google Home – but the new chip could help branch out in much simpler electronics.

 

The team gives IoT devices a potential use case – devices designed to go months on end without charging or changing batteries.  The technology features a “voice activity detection” circuit capable of distinguishing ambient noise from speech, turning on on-board speech recognition hardware when it detects the latter.As the developers say, the new chip will be able to save 90 to 99 percent of energy compared to typical solutions used in modern smartphones for speech recognition. For example, a device used in the IPhone smartphones, which is responsible for communicating with the voice of Siri assistant, spends about one watt of power, while the new chip can perform the same job by consuming 0.2 to 10 milliwatts, depending on the latitude recognizable vocabulary .

 

The chip is essentially designed to be always on in a low-power mode, switching over when voice is detected, thus making it ideal for technologies like wearable devices, which can benefit from speech control, but are required to last much longer on a single charge than a standard handset.

 

naked-science.ru

    Copyright © İnformasiya Texnologiyaları İnstitutu, 2023