Figure 5.
(a) shows the precision (2SD) of each possible combination of algorithm and input signal, with lower values indicating better performance; (b) shows the percentage of combinations which used XA1, XA4, EF3, or none of these techniques (bars may exceed 100% since an algorithm can use more than one of the techniques); (c) shows the percentage of combinations which used frequency- or time-domain RR estimation techniques.
