Open aliman80 opened 2 years ago
@aliman80 Thank you for your interest in our project. It is possible to use our dataset for training object detection algorithms. One would be able to obtain the minimum and maximum coordinate values for the X and Y coordinates of any given object mask/polygon in order to extract bounding boxes. It would be easier with COCO format, as for VOC format an additional connected component analysis step may be necessary.
Hi, thanks for your excellent work, i just want to use Faster RCNN on this dataset for detection task can you tell me how to convert the annotations file to relevant format as currently these are provided in VOC Format/ yolo format