looking to get an easy to install and correct calculation of mAP working, as well as the internal funcs for prediction/groundtruth matching so that we can calculate an object detection confusion matrix and submit this as a metric to torchmetrics (or just for our use depending on time/effort).
we can evaluate, now need to convert groundtruth annotations to pred_dict format
looking to get an easy to install and correct calculation of mAP working, as well as the internal funcs for prediction/groundtruth matching so that we can calculate an object detection confusion matrix and submit this as a metric to torchmetrics (or just for our use depending on time/effort).
we can evaluate, now need to convert groundtruth annotations to pred_dict format