JP-WO2025033195-A5 - Detection device, learning device, detection method, and program

Dates

Publication Date

20260511

Application Date

20240725

Description

The learning synthesis unit 24 may generate one learning synthesis image from one set of learning captured images and learning two-dimensional images. In this case, the blend ratio may be predetermined. In addition, the learning synthesis unit 24 may generate training composite images with different blend ratios for each region within the image. For example, the learning synthesis unit 24 may vary the blend ratio for each object region detected using widely known object detection techniques or semantic segmentation. This is preferable because it allows for the generation of numerous training composite images from a set of training captured images and training 2D images. It is also preferable because it allows for learning of training composite images with various blend ratios. Figure 7 is a block diagram illustrating the hardware configuration of the detection device 10. As shown in Figure 7, the detection device 10 includes a processor 1A, memory 2A, input/output interface 3A, peripheral circuitry 4A, and bus 5A. The peripheral circuitry 4A includes various modules. The detection device 10 does not necessarily have peripheral circuitry 4A. The detection device 10 may also be composed of multiple physically and/or logically separated devices. In this case, each of the multiple devices may have the above hardware configuration.