‘Siri Speech Study’ app to improve Apple speech recognition
Apple launched a new iOS app this month, ‘Siri Speech Study’, designed to collect speech data and improve the speech recognition capabilities of its virtual assistant Siri, reports TechCrunch.
A product of a research study, the app allows participants who have opted in to share voice requests and other feedback with Apple, though documentation does not elaborate on the study’s specific goals. Apple have meanwhile said that the app is being used for Siri improvements via a focus group-like study using human feedback. For example, if Siri misheard a question, users could explain what they were trying to ask, or if Siri on Apple’s HomePod misidentified the speaker in a multi-person household, the participant could note that as well.
Speech pattern recognition and voice-based computing have proliferated in recent years, yet virtual assistants can often still misunderstand types of speech.
The tech giant has been involved in several disputes in recent years over the collection and retention of users’ biometric data, in 2019, an Apple whistleblower shed light on Apple’s collection of confidential consumer voice recordings for manual grading and review, originally reported by The Guardian. Other issues have arisen around consent for the collection of such data such as with Biometric Information Privacy Act (BIPA) of Illinois last year .
Siri Speech Study is currently available in the U.S., Canada, Germany, France, Hong Kong, and India among others, however is only accessible via a direct link.