training on custom dataset

[liuliu66]

trafficstars

I tried your code trained on my own dataset. But I found that your annotation format is .txt which is not standard voc format. I only have the annotations on my dataset of voc format (.xml files) or coco format (.json files). How could I do to use them for training on my dataset?

[zhaoweicai]

I have examples to generate the .txt files for both voc format (.xml files) and coco format (.json files), under $CASCADE_ROOT/data/. Please have a look at them.

[PacteraKun]

@zhaoweicai Hi, Zhaoweicai, I cannot find any file to generate the .txt files for both voc format (.xml files) and coco format (.json files) under $CASCADE_ROOT/data/. Could you please upload again? thanks. I have already found it. thanks.