Figure 5

Breathing-sound classification by human listeners with exactly the same stimuli and experimental procedure as used in the vampire experiment. Results are shown in the same format as in Fig. 3. Error bars represent across-listeners standard errors. While the human listeners were well capable of recognizing breathing sounds recorded in the same session or in a different session in the same physical state, the naïve human listeners failed to recognize the breathing sounds recorded under physical strain (a). Even after the listeners were instructed to ignore breathing-frequency information (b), overall performance did not improve for this most difficult experimental condition.