Provided is an idiosyncrasy sensing system 1 whereby: an idiosyncrasy 13 having an arbitrary characteristic is extracted by an idiosyncrasy extraction unit 3 from a captured image 12 of a subject 11 captured with an image capture unit 2; an idiosyncrasy image 17 having an arbitrary size is clipped out by an idiosyncrasy image clipping unit 4 from the captured image 12 such that the idiosyncrasy 13 overlaps with a center C of the idiosyncrasy image 17; and a classification for the idiosyncrasy 13 is identified by means of machine learning by an identification unit 5, using the idiosyncrasy image 17 as an input.
展开▼