A Nigerian scientist, Dr Aliyu Abubakar, has discovered a better way of identifying individuals with a recent breakthrough study showing that a vision transformer can identify identical twins with remarkable accuracy.
The development has important implications for biometric systems, opening the door to a new way of identifying individuals.
LEADERSHIP recalls that identical twins have always posed a challenge for biometric systems due to their similar genetic material and physical features, making it difficult for traditional systems to distinguish between them.
However, Abubakar’s innovative use of a vision transformer has transformed this process.
The vision transformer is a deep learning model that processes visual information. In the study, Abubakar said thousands of images of human faces were fed into the model, which was then trained to recognize the subtle differences between them.
The results showed that the vision transformer was able to identify identical twins with remarkable accuracy, even when they were dressed similarly and showed no significant differences in their physical appearances.
“The findings of this study are extremely promising and could have far-reaching implications for a variety of applications, including criminal investigations, border control, human resources, and forensic science.
“For example, in human resources, it could help companies to avoid hiring duplicate candidates, and in forensic science, it could assist in investigations where twins are involved, “ Abubakar said.
The study has already drawn attention from the scientific community, with researchers expressing excitement about the potential applications of the technology.
Abubakar is now expanding his research to include other biometric markers, such as iris scans and fingerprints, to develop a more comprehensive system for identifying individuals.
He said the use of a vision transformer to identify identical twins is a major breakthrough in the field of biometrics, with its high accuracy and potential for wide-ranging applications.
