Groundbreaking new tech could offer the gift of speech to people with voice disorders. Chinese scientist Dr Jun Chen developed the ‘voice patch’ with his team at the UCLA Samueli School of Engineering. The device uses artificial intelligence to convert muscle movement into sound.
#AI #tech #biotech #science #China #US
#AI #tech #biotech #science #China #US
Category
🗞
NewsTranscript
00:00 Could you give me your loudest voice and say yeah
00:02 Yeah
00:06 Everly Paddock has spent her entire life trying to find her voice
00:13 Everly was born nearly four months premature
00:16 She couldn't breathe on her own so doctors inserted special tubes in her throat
00:22 Those tubes kept her alive, but they may have damaged her vocal cords in the process
00:30 Today Everly can speak but her voice is little more than a whisper
00:35 Her parents were devastated when they got the news. I do remember the the initial feeling was
00:41 Very just broken-hearted and and you know every every time it felt like okay once we get to this
00:49 yeah, you know this part will will kind of normalize and
00:52 It just felt like one more thing that you know
00:56 We ran up against now there may be hope on the horizon
00:59 This device translates muscle movement into speech
01:04 Here's how it works
01:06 When we speak we create sound waves using our vocal cords or folds
01:11 But we also use certain muscles the voice patch measures just one square inch and sticks to the skin
01:20 There's a thin silicon compound which picks up movement
01:24 Two sets of copper coil are attached to the silicon and in the middle. There's a layer of
01:30 Micromagnets these tiny magnets create a magnetic field each muscle movement
01:37 disturbs that field in a unique way
01:40 That produces an electric signal that passes through the coils
01:44 The voice patch then uses artificial intelligence
01:49 Matching each signal with a word we just use the machine learning
01:53 To extract the features of those the actual signal then we can then actually
01:59 We have a personalized a database
02:02 Actually, then we can power the general electricity with a database
02:08 We contract previously then we can know what words you wanna
02:12 Speak out it will take at least three years before it's on the market
02:17 But Eveli's dad can't wait to see and hear what's possible
02:22 It would be a game changer for for people that can't vocalize. I can't even
02:27 Begin to describe the daily
02:31 Interactions the daily activities that that it would have an impact on lab tests suggest the AI finds the right word
02:38 95% of the time
02:41 There's a slight delay between muscle movement and speech
02:45 But dr. Chen says it could soon disappear
02:48 Eventually, we will not feel any delay. Just like I talk to you now. I'm using my vocal fold
02:55 You can you cannot feel any delay yet, but that is the ultimate goal
02:59 We are still working toward that direction Eveli's journey has been tough
03:04 But if everything goes well, this technology could change her life
03:08 Giving her the chance to share her voice with the world
03:13 Michael Merilia CGTN