Conceptual Clarification fro FP vs FN


I want to verify my understanding.

Screenshot 2023-12-29 at 1.34.03 AM

Specifically, in Figure 7 of the report, one pair of a Ground Truth and a Participant node is both a FP and a FN. Is this always the case for type 2 (no ground truth nodes within tolerance around participant node) FP’s?



That’s not always the case. A FN represents a missed ground truth node, without any participant node around the tolerance interval. The blue FP you see in the image represents an unmatched participant node due to it is outside of the tolerance interval of any ground truth node. In case the participant wouldn’t have predicted anything at all there, you would have a FN but not a FP.