I'm a last year PhD student under the supervision of Daniel Gatica-Perez at the Ecole Polytechnique Fédérale de Lausanne (EPFL) and research assistant at Idiap Research Institute.

My research area of interest is the analysis of sensor data both offline and online with the purpose of understanding human social behavior and communication, as well as to explore how people engage and experiment with multimedia content. In this line, my doctoral research is focused on the analysis of human communication, interaction, and multimedia engagement with online social video in sites such as YouTube, addressing the automatic analysis of conversational video blogs (vlogs).

In an early piece of work, I addressed the analysis of the aggregated behavior analysis of YouTube users based social network metadata, investigating on the different patterns of participation and interaction of YouTube users from the lense of self-assigned user categories.

I then focused on the investigation of a particular type of social video, which are conversational vlogs. This setting represents the simplest vlogging scenario and the one that features most conversational behavior, compared to other vlogging styles. Specifically, Daniel and I addressed the problem of automatic conversational analysis of vlogs (VlogSense), which can be considered as a new research domain in social media.

In the context of vlogsense, we investigated the problem of vloggers' personaliy impression formation and prediction using crowdsourced annotations and automatic content analysis