Table 1. Average precision and recall results for bat search-phase call detection algorithms across three different test sets iBats Romania and Bulgaria; iBats UK; and Norfolk Bat Survey.

Detection Algorithms
				BatDetect
Average Precision	SonoBat	SCAN’R	Kaleidoscope	Segment	Random Forest	CNNFAST	CNNFULL
iBats (R&B)	0.265	0.239	0.189	0.299	0.674	0.863	0.895
iBats (UK)	0.200	0.142	0.144	0.324	0.648	0.781	0.866
NBP (Norfolk)	0.473	0.456	0.553	0.506	0.630	0.861	0.882
Recall at 0.95
iBats (R&B)	0	0.251	0	0	0.568	0.777	0.818
iBats (UK)	0	0	0	0	0.324	0.570	0.670
NBP (Norfolk)	0.184	0.470	0	0	0.049	0.781	0.754

Large numbers indicate better performance. Recall results are reported at 0.95 precision, where zero indicates that the detector algorithm was unable to achieve a precision greater than 0.95 at any recall level. The results for the best performing algorithm are underlined. Details of the test datasets and detection algorithms are given in the text.