US 12,267,570 B2
Machine learning device, image processing device, machine learning method, and machine learning program
Yincheng Yang, Yokohama (JP); Shingo Kida, Yokohama (JP); and Hideki Takehara, Yokohama (JP)
Assigned to JVCKENWOOD Corporation, Yokohama (JP)
Filed by JVCKENWOOD Corporation, Yokohama (JP)
Filed on Feb. 24, 2023, as Appl. No. 18/173,819.
Application 18/173,819 is a continuation of application No. PCT/JP2021/003568, filed on Feb. 1, 2021.
Claims priority of application No. 2020-142707 (JP), filed on Aug. 26, 2020.
Prior Publication US 2023/0199281 A1, Jun. 22, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. H04N 23/11 (2023.01); G06V 10/70 (2022.01)
CPC H04N 23/11 (2023.01) [G06V 10/70 (2022.01)] 5 Claims
OG exemplary drawing
 
1. A machine learning device comprising:
a far-infrared image training data acquisition unit that acquires a far-infrared image in a first predetermined time zone and a far-infrared image in a second predetermined time zone;
a visible light image training data acquisition unit that acquires a visible light image in the second predetermined time zone; and
a visible light image generation model learning unit that generates a trained visible light image generation model that generates the visible light image in the second predetermined time zone from the far-infrared image in the first predetermined time zone,
wherein the visible light image generation model learning unit includes:
a first generation model learning unit that machine-learns the far-infrared image in the first predetermined time zone and the far-infrared image in the second predetermined time zone as training data, and generates a trained first generation model that generates the far-infrared image in the second predetermined time zone from the far-infrared image in the first predetermined time zone, and
a second generation model learning unit that machine-learns the far-infrared image in the second predetermined time zone and the visible light image in the second predetermined time zone as training data, and generates a trained second generation model that generates the visible light image in the second predetermined time zone from the far-infrared image in the second predetermined time zone.