This technology uses human breath for identification, achieving an accuracy of up to 97%.
Throughout history, humans have experimented with various technologies for identification and security on smartphones, such as passwords, patterns, fingerprints, and facial recognition. However, none of these methods have provided perfect security and functionality in real-world applications. Consequently, scientists have developed an entirely new identification technology: breath recognition.
This is a new biometric technology beyond facial recognition and fingerprints… (Image: Zenger).
Specifically, researchers from Kyushu University and the University of Tokyo have unveiled a new method for unlocking smartphones. They claim that users only need to breathe to unlock their phones. Moreover, the accuracy of this technology reached 97% in the first test.
Regarding the benefits of breath recognition compared to traditional fingerprint methods, Chaiyanut Jirayupat, the lead author of the study, stated that this technology is developed based on the unique characteristics of each individual but does not present them too overtly.
“Physical characteristics like fingerprints and facial features can be copied or altered if injured,” he noted. Therefore, the scent of a person becomes a new biometric identification trait. Devices only need to identify the “type of gas” produced by human skin.
According to Gizchina, this method employs artificial intelligence (AI) technology known as “electronic nose” along with an olfactory sensor system. This allows the technology to analyze scents in the air and accurately determine the components present in those scents.
Users can unlock their phones just by breathing. (Image: Gizchina).
In relation to the human body, our breath is composed of many compounds. Every time we eat, our breath changes according to the food consumed. Therefore, each person has a unique chemical signature in their breath. This chemical signature can even be used to identify certain diseases like diabetes.
The research from Kyushu University and the University of Tokyo identified at least 28 compounds present in human breath by utilizing various olfactory sensors. Scientists also employed machine learning technology to analyze the chemical components in the breath, subsequently identifying each individual’s identity.
Surprisingly, the accuracy rate reached 97.8%. However, this figure is still not sufficient to prove that breath recognition will replace other technologies. Specifically, facial recognition achieves an accuracy of up to 99.97%, while fingerprints reach 98.6%.
Additionally, this method has its limitations. “Human skin typically does not produce enough gas for devices to recognize,” researcher Chaiyanut Jirayupat stated.
Olfactory sensors help devices recognize user breath. (Image: Zenger).
According to Gizchina, unlocking phones with breath is not an overly strange method. Previously, many biometric identification technologies have utilized other human characteristics.
For instance, Amazon has used palm scanning for payment verification on Amazon One. Mastercard has also developed a new identification method based on individual gait. Additionally, iris recognition, DNA, fingerprinting, and finger vein patterns have all been incorporated into identity verification technologies.
- However, all these methods have not proven to be entirely effective as these characteristics can change. They are all physical traits that can deform if injured. Even fingerprint recognition, regarded as the most secure method, is not perfect, Gizchina noted.
- American startup drone sets record lifting 374kg cargo
- Concept of nuclear-powered aircraft with a speed of 4,000km/h
- Development of tiny robots capable of transforming to deliver medicine within the human body
