Apple is training Siri to better understand users who stutter

Apple is looking into how to improve Siri so that it can better understand the needs of users who stutter, according to new details shared by The Wall Street Journal in an article about how the company trains the voice assistant to process the voices of users with speech impairments.

Apple has built a database of 28,000 audio clips provided by users who stutter. According to an Apple spokesperson, this database will be used to train Siri and improve the voice recognition system's ability to understand users with speech impairments.

In addition to improving Siri's ability to understand users with speech impairments, Apple has also added a Hold to Talk feature to Siri, which allows users to freely control the time Siri listens to voice, thus preventing Siri from interrupting users with stuttering.

With the Type to Siri feature first introduced in iOS 11, users can use Siri without using voice.

We've learned that Apple plans to release a research report this week on its work to improve Siri, which will provide more details about the company's work in this area.

Google and Amazon are also reportedly working to train Google Assistant and Alexa to better understand the voices of users with speech impairments. Google is collecting voice data from users with speech impairments, and Amazon launched the Alexa Fund in December to build an algorithm to recognize the unique vocal patterns of users with speech impairments.