[15p-M_123-1~8] FS.1 Focused Session "AI Electronics"

2:15 PM - 2:30 PM JST(5:15 AM - 5:30 AM UTC)

[15p-M_123-2] Quantum anomaly detection inside fruit using external images

〇Takao Tomono¹, Kazuya Tsujimura² (1.Keio Univ., 2.TOPPAN Holdinga)

Keywords:

quantum machine learning,anomaly detection

In this study, we demonstrated the discriminative capability of a support vector machine (SVM) using quantum kernels for anomaly detection with a small dataset. A total of 66 apples (33 normal and 33 defective) were used. Apples with internal browning were imaged under LED illumination, and after binarization, features were extracted using principal component analysis (PCA). The classification performance was evaluated using both a quantum simulator and a quantum computer. The results indicate that an effective learning model can be constructed even with a small dataset, and furthermore suggest that only the first to third principal components significantly contribute to the discrimination performance.