| CPC G06V 10/7747 (2022.01) [G06V 10/95 (2022.01); G06V 20/56 (2022.01)] | 4 Claims |
|
1. An information processing device, comprising a processor configured to execute program instructions to:
acquire a plurality of captured images of an outside of a vehicle;
detect, from the acquired plurality of captured images, a plurality of images related to unsafe driving using a learned model;
extract, from the detected plurality of images related to unsafe driving, an image to be a candidate for relearning data of the learned model; and
conduct relearning of the learned model using, as relearning data, an image that has been identified not as an image related to unsafe driving, among images that each have been extracted because a confidence thereof was equal to or greater than a threshold and a predetermined condition was satisfied;
wherein, when a confidence of a detected image related to unsafe driving is less than the threshold, or when the confidence of the detected image related to unsafe driving is equal to or greater than the threshold and the detected image related to unsafe driving satisfies the predetermined condition under which false detection is likely to occur, extract the detected image related to unsafe driving.
|