An Artificial Intelligence (AI) approach for studying dolphin communication

An Artificial Intelligence (AI) approach for studying dolphin communication

By In PhD proposals 2018 On February 13, 2019

Project: An Artificial Intelligence (AI) approach for studying dolphin communication

Laboratory: Neuronal circuit dynamics and behaviour

Affiliation: Institut de Biologie de l’Ecole Normale Superieure
Address: 46 rue d’ulm 75005 paris

LAB Director
Name: German SUMBRE
Phone number: 0144322367

Name: German Sumbre
Phone number: 0144322367

Subject Keywords: dolphins, communication, language, artificial intelligence, behaviour
Summary of lab’s interests: The multidisciplinary approach will be a collaboration of 3 institutes. Data acquisition at a dolphin reef in Israel, data analysis of the data using methods from neuronal circuit analysis at the ecole normale superieure, Paris and data analysis using methods from artificial intelligence at the champalimaud institute in portugal
Project summary: An Artificial Intelligence (AI) approach for studying dolphin communication
A collaboration between the labs of Germán Sumbre, IBENS, Gonzalo de Polavieja, Champalimaud, Portugal, and Dolphin Reef, Israel.
Dolphins use vocal signals to communicate via two different types of sounds: pure tones and pulsed sounds. Pure tones are frequency modulated sounds such as whistles, chirps and screams. Dolphins produce whistles in social contexts. For example, individual dolphins have their own “signature whistle” or “names” that are produced by other members of the group. Pulsed sounds are clicks generated at regular brief intervals, and are used for echolocation. These clicks can be sometimes generated at high frequencies (short intervals), and are thought to carry information about emotional states.
For this project, we intend to record dolphin sounds and monitor their behaviour in a quasi-natural environment. Using an AI approach, we intend to find correlations between specific sound properties and certain types of behaviours to learn basic principles underlying dolphin communication.
As part of the proposed internship project, the student will work within the framework of a collaboration between the labs of German Sumbre (Neurosciences, IBENS), Gonzalo de Polavieja (Champalimaud, Portugal) and the Dolphin reef (Israel).
The student will be supervised by German Sumbre, but he/she will spend some time at the de Polavieja lab to learn techniques on AI. A potential visit to the dolphin reef to acquire new datasets may be conceivable.
Interdisciplinary aspect of the project: The project will combine the data acquisition of behaviour and sound of dolphins and computational methods from data mining and artificial intelligence for classifying the emitted sound and correlate these to the motor and social behaviours.