Scripts for kinds of converting dataset labels.
- Partly referred to CasiaFan/Dataset_to_VOC_converter
- cytoolz
- lxml
$ python3 voc2yolo.py --img_dir <image directory> --anno_dir <annotation directory> --out_dir <output directory>
By default, this script will automatically parse all annotation files to get class list, sort and output a name file when not given classes are not given. Manually add class in code(line 13) will skip this parsing stage and use the given classes.
More usage details see:
$ python3 voc2yolo.py --help