How to detect specific class in image from label.txt in object detection?

Actually I have 50 images and its respective labels.txt but in each label file I have multiple other classes which I don’t want to detect. Here I have attached a sample image and its label.txt to get better understanding

[url]https://github.com/dusty-nv/jetson-inference/files/3725047/000000001455.txt[/url]

Here I just want to detect cardboard in image and cardboard class object from label.txt. So I am just confused what to put in the custom classes option while preparing the dataset. I kept it empty for the first time and for second time I just put the class cardboard which I basically want to get detected but my model didn’t get trained. map showed 0. can anyone take a look? thanks cheers. Any kind of advice, suggestions would be really helpful.