MACHINE LEARNING DEVICE

• Main Authors: MURATA YOSHITAKA, HIROIKE ATSUSHI, IMADA TSUTOMU, MORITA KENICHI, WATANABE YUKI
• Format: Patent
• Language: Chinese; English
• Published: 21.12.2018
• Subjects: CALCULATING; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The purpose of the present invention is to provide a machine learning device that is capable of reliably and quickly improving the accuracy of image identification. This machine learning device 1 is provided with: an image database 422 for storing a plurality of images and the image characteristic quantity of these images; and a processor 41 that is connected to the image database 422 and that performs machine learning using the plurality of images and the image characteristic quantity stored in the image database 422. The processor 41 preferentially selects as images to be used for machine learning a prescribed number of images, from among those stored in the image database 422, which are images other than the images used in previous machine learning and which have little resemblance to the images used in previous machine learning, and then uses these selected images to perform new machine learning. 本发明目的在于提供种能够可靠且快速提高图像识别精度的机器学习装置。本发明的机器学习装置(1)包括：图像数据库(422)，其存储多个图像和该多个图像的图像特征量；以及处理器(41)，其与该图像数据库(422)连接，使用在图