Every sound we hear has a unique signature thanks to the way it was created and which objects the sound waves have passed through. A team of South Korean researchers are now exploring whether the unique bioacoustic signatures created as sound waves pass through humans can be used to identify individuals. Their work, described in a study published 4 October in IEEE Transactions on Cybernetics, suggests this technique can identify a person with 97 percent accuracy.
The biometric system developed by the group at Electronics and Telecommunications Research Institute (ETRI) uses a transducer to generate vibrations and thus sound waves, which pass through a given body part on a person. In this a case, a finger is easily accessible and convenient. After the sound has passed through the skin, bones, and other tissues, a sensor picks up the unique bioacoustic signature. Teasing apart the distinct signatures of individuals is further boosted using modeling.
“Modeling allowed us to infer what structures or material features of the human body actually differentiated people,” explains Joo Yong Sim, one of the ETRI researchers who conducted the study. “For example, we could see how the structure, size, and weight of the bones, as well as the stiffness of the joints, affect the bioacoustics spectrum.”
The approach is effective enough to distinguish different fingers on the same hand. This means that a person must use the same finger that was originally analyzed for authentication.
Notably, the researchers were concerned that the accuracy of this approach could diminish with time, since the human body constantly changes its cells, matrices, and fluid content. To account for this, they acquired the acoustic data of participants at three separate intervals, each 30 days apart.
Conceptual diagram of bioacoustics identity authenticationIllustration: Joo Yong Sim/ETRI
“We were very surprised that people's bioacoustics spectral pattern maintained well over time, despite the concern that the pattern would change greatly,” says Sim. “These results suggest that the bioacoustics signature reflects more anatomical features than changes in water, body temperature, or biomolecule concentration in blood that change from day to day.”
While measuring changes in acoustic vibrations is fairly accurate, it does not yet match the accuracy of fingerprints or iris scans. In the future, Sim’s team plans to add additional sensors to the system to boost its accuracy. In the meantime, they are exploring ways to commercialize the technology in phones and wearable devices.
What’s more, development of this technology has come with an unexpected twist. The technique is so accurate at analyzing tissues, the team is now exploring ways to use it for diagnosing musculoskeletal disease.
An abridged version of this post appears in the December 2019 print issue as “It’s Personal: Bioacoustic Identification.”