MIT develops low-power voice recognition chip

IN response to the massive anticipation regarding the voice-controlled electronics, researchers at the Massachusetts Institute of Technology (MIT) have built a low power chip for automatic voice recognition. The chip developed by MIT, requires between 0.2 and 10 milliwatts of power to run, depending highly on the amount of words it has to recognize per speech. The speech recognition chips in cell phones on the other hand require around 1 watt of power. The power used by the chip developed by MIT saves around 90 per cent to 99 per cent of electricity, thus making the voice control option available for simpler devices like vehicles, appliances and manufacturing equipment.

Even devices from the technological zone or more commonly known as IoT, will very soon be equipped with the voice recognition option, which will allow users to directly report the information to the desired servers in a coordinated manner. Per Anantha Chandrakasan, the professor of Electrical Engineering and Computer Science at MIT, the concept of speech input will very soon become a common interface for many wearable and intelligent devices. According to Michael Price, who executed the designing, this chip was not developed for any particular application. A paper on the chip was presented at the International Solid-State Circuits Conference by Price, Chandrakasan, and Jim Glass, a senior research scientist of MIT’s Computer Science and AI Laboratory.