Image segmentation to Voc

I’ve exported my image/label to export.json file and run the voc_converter tool. It seems like the voc_converter tool only focuses on object detection? I’d expect it to create the masks for segmentation. Am I missing something?