Vai al contenuto principale

Filippo Carugati

  • Phd: 37th cycle
  • Matriculation number: 851254

Contacts

Supervisor

Marco Gamba

Phd thesis

Co-occurence of facial gestures and vocalizations across primates

upload_indri-skeleton_low.jpg

As humans, also primates use multiple communication channels to convey signals – that can be olfactory, vocal, and visual (declined as body and facial gestures). These signals can be simultaneously or sequentially integrated. The resulting signals are “multimodal” since they are generated and perceived by the integration of different communicative channels. The rising interest in primate multimodal communication highlighted how the methodologies applied to the study of each communicative channel are not easily comparable to each other.

Among the different communicative channels, facial gestures have been generally studied through the application of discrete labeling, focused primarily on the ethological context of each facial display or on its emotional valence. Despite an increase in experimental studies about facial expressions during the last years, most of the current knowledge derives from descriptive analysis, based on observational and operator-controlled data collections, which generally results time-consuming and prone to human error or subjective interpretations. The current state-of-the-art about primate facial gestures suggests the need for a tool based on accurate automatic detection. Promising results derived from the application of artificial intelligence for the study of animal behavior can also be extended to different communicative modalities, such as acoustic signals or visual signals.

My Ph.D. project aims to develop a new methodology, based on deep learning algorithms, to extract quantitative parameters about primate facial gestures and to apply this technique for the estimation of the co-occurrence rate between facial gestures and vocalization among different primate species. I believe that this technology has the potential to pave the path for a new methodological approach rooted in the automatic recognition of primate faces, making also possible a major integration of data derived from different communicative modalities.

Last update: 17/11/2022 13:15
Location: https://dott-sbba.campusnet.unito.it/robots.html
Non cliccare qui!