From ShutEye to SleepScore, several smartphone apps available if you’re trying to better understand how snoring affects your rest, allows you to leave the microphone on overnight to capture grunts. in the nose and rumbling throat reverberation. But while smartphone apps are useful for tracking presence of snoring, their accuracy remains an issue when applied to real-world bedrooms that have outside noise and are audible to many people.
Preliminary study from the University of Southampton to see if your snoring sounds characteristic sound which can be used for identification. “How exactly do you track snoring or coughing?” asked Jagmohan Chauhan, an assistant professor at the university, who did the research. Machine learning models, specifically deep neural networkcan assist in verifying who is performing that snoring symphony.
Although this research is relatively young, it is built peer-reviewed research used machine learning to verify the creators of another data-rich sound, often heard through the optimistic silence of the night: the cough.
Researchers from Google and the University of Washington mixed the sounds of human voices and coughs into a dataset, then used multitasking learning to verify who made a particular cough in record. IN their researchThe AI performed 10% better than human reviewers at identifying who coughed out of a small group of people.
Matt Whitehill, a graduate student who worked on the paper that identified cough, questioned some of the research methods on snoring and suggested that more rigorous testing would reduce its effectiveness. However, he finds the broader concept of sound recognition valid. “We’ve shown that you can do it with coughs. Looks like you can do the same thing with snoring,” says Whitehill.
This audio-based AI segment isn’t as widely covered (and certainly not in rhetorical terms) as natural language processors like OpenAI’s ChatGPT. But despite that, some companies are looking to use AI to analyze audio recordings and improve your health.
Resonance, a Swiss company focused on AI-powered lung disease symptom detection, has released medical software that is CE certified and available to Swiss people through the myCough app. While the software is not designed to diagnose illnesses, it can help users keep track of how many times they cough during the night and which types of coughs are most common. This provides users with a more complete understanding of their cough patterns while they decide whether to seek medical advice.
David Cleres, co-founder and chief technology officer at Resmonics, sees the potential of deep learning techniques to identify a specific person’s cough or snoring, but believes major breakthroughs are still needed for the area of research. save this AI. “We learned the hard way at Resmonics that endurance to change in devices and recording locations is just as difficult to achieve as resistance to variations from different groups of people,” Cleres wrote via email. different use. Not only is it hard to find a dataset with lots of natural cough and snoring recordings, but it’s also hard to predict the microphone quality of a 5-year-old iPhone and where someone would choose to leave it at night.
So the sounds you make in bed at night can be tracked by the AI and are different from the night sounds made by other people in your household. Snoring can also be used as a biometric linked to you, like a fingerprint? More research is needed before any premature conclusions can be reached. “If you look at it from a health perspective, it can work,” says Chauhan. “From a biometric perspective, we can’t be sure.” Jagmohan is also interested in discovering how Signal Processingwithout the help of machine learning models, can be used to assist in detecting snorers.
When it comes AI in the healthcare environment, eager researchers and intrepid entrepreneurs continue to run into the same problem: the lack of available quality data. The lack of diverse data for AI training can be a tangible danger to patients. For example, an algorithm used in hospitals in the US give up care priority of black patients. Without robust data sets and thoughtful modeling, AI often behaves differently in real-world situations than in hygienic practice settings.
“People are really turning to deep neural networks,” says Whitehill. This data-intensive approach further increases the need for bulk recordings to produce quality cough and snoring research. A machine learning model that tracks when you snore or have trouble breathing isn’t as easy to remember as one chatbots created existential sonnets about Crunchwrap Supreme by Taco Bell. It is still worth pursuing vigorously. While innovative AI remains a top priority for many in Silicon Valley, it would be a mistake to hit the snooze button on other AI apps and ignore their vibrant capabilities.
