Inquiry on Customizing Classification Labels in ObjectDetectionImage

Hi, all
I am currently using ObjectDetectionImage tool for object detection in my project. However, I am encountering an issue with the classification labels in the output. The labels, such as “car” and “motorbike,” do not match my custom categories. I would like to know how to configure or modify the tool so that it uses my own category names for detection.

Below is an example of the output I receive:

image data size: w:1920, h:1080, c:3

YOLOv7t_SD tick[0]

[image 30] test\20250328-9-19.jpg

SD_file_load_file: sd:/test\20250328-9-19.jpg

file_size: 118982 bytes.

image_id = 02

object num = 3

save file to sd:/test\20250328-9-02.json

Image: test\20250328-9-02, OBJECT IDENTIFIED: car, with a probability of: 0.243

Image: test\20250328-9-02, OBJECT IDENTIFIED: motorbike, with a probability of: 0.380

Image: test\20250328-9-02, OBJECT IDENTIFIED: motorbike, with a probability of: 0.243

How can I modify the classification labels to reflect my custom categories?
Best regards

Hi @Clay , you may modify the classification labels in NNObjectDetectionImage.cpp on char NNObjectDetectionImage::coco_name[80][20] in line 24. Thank you.