作者: Eugenia V. Bragina , Irina R. Beme
DOI: 10.1007/S10211-010-0073-6
关键词: Statistical classification 、 Artificial intelligence 、 Spectrogram 、 Vocal communication 、 Method of analysis 、 Communication 、 Male to female 、 Pattern recognition 、 Grus leucogeranus 、 Biology 、 Population 、 Linear discriminant analysis
摘要: Vocal individuality varies between species and/or ontogenesis stages depending on needs in the vocal recognition, but also estimation of individual differences depends method analysis. We studied pair-specific duets elicited by mating pairs Siberian crane Grus leucogeranus. quantitatively described duet structure and compared visual statistical classification methods pair identification duet. Three were used: discriminant analysis, trees spectrogram. found significant interpair differences. The differ that is ratio male- female-initiated number male to female calls; temporal-frequency characteristics are pair-specific, too. All showed high differences, which exceeded random values significantly. Discriminant analysis stepwise procedure based 11 parameters resulted 97.3% correctly assigned duets. Human observers 80.7% spectrograms. Our data provide a basis for remote monitoring this endangered with wild population only 3,000 birds.