New project on elephant communication – Decoding elephant communication with AI funded by the WWTF, Vienna Science and Technology Fund Grand ID 10.47379/LS23024
Elephant research and conservation issues reveal the need for a deeper understanding of their behavior, ecology, cognition, perception, and communication in an increasingly human-dominated environment to ensure their survival and well-being. Elephants are a socially and spatially flexible species and sound is a crucial mode of communication. The most used call type is the low-frequency rumble, with fundamental frequencies within the infrasonic range generated by passive vocal fold vibration, following the same principle of voice production as in humans.
Elephants possess vocal plasticity, allowing them to alter the structure of their calls in response to environmental and biological conditions. Although it is known that elephant calls convey biologically significant information, the diversity of their calls makes it difficult to decode and interpret their communication pattern. AI offers potential to analyze communication patterns that can be used to generate novel hypotheses about signal function and may ultimately aid in the design of controlled experiments to verify these hypotheses via synthesized signals. The central research question investigated in this project is: Can we leverage AI to decode elephant communication from large-scale data?
