Using AI in collective animal behavior without (too many) tears

Event Slider

Several seminars are held weekly at the Instituto Gulbenkian de Ciência, an initiative that aims to bring together all researchers around the topics under discussion.

The sessions, with internal researchers or guests, contribute to stimulate the open and extremely collaborative culture of the IGC.

You can read the abstract of this seminar to learn more about it.

 

I would like to convey the opportunities that AI offers to study animal behavior. I will spend a few minutes giving an intuition of how AI works and then explain how we have used it in the lab to study collective animal behavior. One clear opportunity AI offers is to extract richer information from our experiments, that I will illustrate mainly with a tracking system. Less obvious is perhaps how to use AI to understand animal behavior. I will show how we used it to find a model of collective behavior that is both accurate and transparent. I believe the AI modelling is quite close to what a biologist expects from a model: a quantitative description that does not oversimplify too much the underlying Biology.

 


SPEAKER

Gonzalo Polavieja
Champalimaud Foundation, Portugal

 

HOST
Rui Oliveira

Cookies settings

Cookies Selection

This website uses cookies to improve your browsing experience, security, and its website performance. We may also use cookies to share information on social media and to display messages and advertisements personalised to your interests, both on our website and in others.