[15p-WL1_301-1~9] Application of informaics technology in the field of nanobiotechnology

5:10 PM - 5:45 PM JST(8:10 AM - 8:45 AM UTC)

[15p-WL1_301-8] Odor Molecule Recognition Using AI Olfactory Sensors

〇Genki Yoshikawa¹ (1.NIMS)

Keywords:

sensor,gas,odor

In this talk, I will present recent advances in odor molecule identification based on AI analysis of multidimensional data obtained from olfactory sensors. Using Membrane-type Surface stress Sensors (MSS) as a representative example, the presentation will cover machine-learning-based quantitative estimation, visualization of the identification process using explainable AI (XAI), reconstruction of complex odors based on the concept of quasi-primary odors, and robust feature engineering using transfer function ratios (TFRs). These approaches highlight how carefully designed data representations enhance both performance and interpretability in AI olfactory sensing. Furthermore, recent field demonstrations in agricultural and medical applications will be introduced, along with future perspectives on the integration of olfactory sensing and informatics.