Ground truth detections, with "don't care" areas included, were extracted using:
./results/test_eval_on_ground_truth/convert_gt.py
To compare evaluation results on all ground truth detections (with "don't care" areas included) from the current script with results from the proposed script, execute:
$ ./compare_scripts.sh