Ear acoustic authentication is a biometric authentication technology that recognizes the acoustic characteristics of the ear canal to authenticate users.
However, compared to fingerprints, ear acoustic authentication has not been studied sufficiently with regards to the individuality of the acoustic characteristics of the ear canal.
Therefore, a study on the visualization of ear canal acoustic characteristic differences using t-distributed stochastic neighbor embedding (t-SNE) which expresses the similarity in high-dimensional space and estimates the similarity in low-dimensional space, was conducted.
Authors: Rei Ominato (NIT, Nagaoka College), Shohei Yano (NIT, Nagaoka College), Naoki Wakui (NIT, Nagaoka College) and Shinnosuke Takamichi (The University of Tokyo)