Classification and automatic transcription of primate calls

Area of Science:

• Bioacoustics
• Primatology
• Computational Biology

Background:

• Manual acoustic analysis of primate vocalizations is time-consuming.
• Identifying and labeling primate calls requires specialized expertise.
• Current methods often lack automation and open accessibility.

Purpose of the Study:

• To develop an automated tool for acoustic analysis and transcription of primate calls.
• To provide an openly available system for processing raw field recordings.
• To improve the efficiency and accuracy of primate vocalization analysis.

Main Methods:

• Development of an automated acoustic analysis system.
• Implementation of algorithms for call detection and time-alignment.
• Utilizing raw field recordings as input without manual feature selection.

Main Results:

• The system accurately transcribes primate calls with time-aligned labels.
• A majority of call start times are predicted within 200 milliseconds.
• The tool requires no manual acoustic analysis or spectral feature selection by researchers.

Conclusions:

• The automated tool offers an efficient and accessible solution for primate call analysis.
• This system significantly reduces the manual effort required in bioacoustic studies.
• Open availability promotes wider adoption and advancement in primate vocalization research.