English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Researches on sequential vocalization often require analysis of vocalizationsin long continuous sounds. In such studies as developmental ones or studiesacross generations in which days or months of vocalizations must be analyzed,methods for automatic recognition would be strongly desired. Although methodsfor automatic speech recognition for application purposes have been intensivelystudied, blindly applying them for biological purposes may not be an optimalsolution. This is because, unlike human speech recognition, analysis ofsequential vocalizations often requires accurate extraction of timinginformation. In the present study we propose automated systems suitable forrecognizing birdsong, one of the most intensively investigated sequentialvocalizations, focusing on the three properties of the birdsong. First, a songis a sequence of vocal elements, called notes, which can be grouped intocategories. Second, temporal structure of birdsong is precisely controlled,meaning that temporal information is important in song analysis. Finally, notesare produced according to certain probabilistic rules, which may facilitate theaccurate song recognition. We divided the procedure of song recognition intothree sub-steps: local classification, boundary detection, and globalsequencing, each of which corresponds to each of the three properties ofbirdsong. We compared the performances of several different ways to arrangethese three steps. As results, we demonstrated a hybrid model of a deep neuralnetwork and a hidden Markov model is effective in recognizing birdsong withvariable note sequences. We propose suitable arrangements of methods accordingto whether accurate boundary detection is needed. Also we designed the newmeasure to jointly evaluate the accuracy of note classification and boundarydetection. Our methods should be applicable, with small modification andtuning, to the songs in other species that hold the three properties of thesequential vocalization.

Automatic recognition of element classes and boundaries in the birdsong with variable sequences