Network Associates Successful in WWTF Life Sciences 2023 Call “Understanding Biology with AI/ML”


Proposals by Marisa Hoeschele, Angela Stöger-Horwath, and Manuel Zimmer and Moritz Grosse-Wentrup were amongst the successfuly funded projects in the WWTF Life Sciences 2023 Call “Understanding Biology with AI/ML”

LS23-014 Marisa Hoeschele (ÖAW - Austrian Academy of Sciences) Analysis of Nonhuman Intercommunication with Machine Learning

Most research on animal vocalizations considers units of sound separated by silence, e.g., dog barks. However, humans can create entire sentences without a break, many of which are unique. Similarly, budgerigars (a small parrot species) create unique utterances with components that resemble consonants and vowels. To understand whether the utterances have some meaning resembling language, we need to be able to study budgerigar vocalizations in a natural context. However, because many birds vocalize simultaneously (similar to humans at a party) this has not yet been possible. In the proposed project, we will use recent advances in machine learning and signal processing to extract recordings of each individual bird from multi-microphone and multi-camera recordings of the aviaries. We will then use behavioral experiments to learn about the meaning of the utterances. Ultimately, this project will help us to understand whether humans are the only species on the planet with language.

  • Duration: 48 months, Funding volume: € 799,885

LS23-024 Angela Stöger-Horwath (ÖAW - Austrian Academy of Sciences) Decoding elephant communication with AI

To secure elephant survival in an increasingly human-dominated world, we must first understand their behavior, thinking, and communication patterns. Elephants communicate significantly through vocalizations. They send relevant information via their calls, yet each one is unique, with variances in numerous auditory aspects. Working with massive data sets of vocalizations that people cannot easily evaluate manually will be required, because it is unknown which auditory patterns encode critical information. The main question is whether Artificial Intelligence (AI) can aid in the decoding of elephant communication.
The highest level of verification is provided by combining innovative models based on machine learning and artificial intelligence to decode elephant communication patterns and testing our findings not just in the lab but also on elephants in the wild. We, a group of biologists and computer scientists, will take on the challenge of decoding elephant language. 

  • Duration: 48 months, Funding volume: € 876,188

LS23-070 Manuel Zimmer (University of Vienna) (Co-PI: Moritz Grosse-Wentrup) An interdisciplinary approach to learn and test the causal mapping between neural network dynamics and behavior

Previous research has investigated how brains encode animal behavior and found that it is linked to how large groups of nerve cells work together in tight coordination. We still don't understand how the actions of such large groups of nerve cells is controlled. To solve this, we bring together a team of experts in artificial intelligence, computational biology and neurobiology. We will study a small creature, a worm called C. elegans, which has only 302 nerve cells and can therefore be better studied and understood. For example, it is possible to visualize brain activity including all of its individual nerve cells. We will use new computer techniques to learn from its brain activity and simulate how its nerve cells communicate. Literarily, we will learn how to read its mind! We will then test our ideas by using special genetic tricks to manipulate brain activity with light, in order to control specific groups of nerve cells and then observe how this affects decision making of the worm.

  • Duration: 48 months, Funding volume: € 799,998