GPI Seminar Series: Xavier Alameda

Xavier Alameda

Xavier Alameda, Post-doctoral fellow, MHUG/Trento
Detecting social attention attractors in free-standing conversational groups through multimodal head and body pose estimation
Thursday December 3rd, 10h, Room D5-007

During natural social gatherings, humans tend to organize themselves in the so-called free-standing conversational groups (FCGs). Studying FCGs in unstructured social settings (e.g., cocktail party ) is gratifying due to the wealth of information available at the group (mining social networks) and individual (recognizing native behavioural and personality traits) levels. However, analysing social scenes involving FCGs is also highly challenging due to the difficulty in extracting behavioural cues such as target locations, their speaking activity and head/body pose due to crowdedness and presence of extreme occlusions. Importantly, visual information typically obtained with a distributed camera network might not suffice to achieve the sought robustness. In this line of thought, recent advances in wearable sensing technology open the door to multimodal and richer information flows. In this study we cast the head and body pose estimation problem into a matrix completion task. We introduce a framework able to fuse multimodal data emanating from a combination of distributed and wearable sensors, taking into account the temporal consistency, the head/body coupling and the noise inherent to the scenario. We report results on the novel and challenging SALSA dataset, containing visual, auditory and infra-red recordings of 18 people interacting in a regular indoor environment. We demonstrate the soundness of the proposed method and the usability for higher-level tasks such as the detection of F-formations and the discovery of social attention attractors.

Short Biography:
I obtained a degree of Mathematics from the Faculty of Mathematics and Statistics in 2008 and a Degree in Telecommunications Engineering from the School of Telecommunications in 2009 both from the Technical University of Catalonia — BarcelonaTech. Afterward I went to Grenoble to follow the MoSIG masters program in Graphics, Vision and Robotics, offered at UFR IM2AG from University Joseph Fourier (Grenoble 1) and Grenoble INP – Ensimag (formerly ENSIMAG, INP Grenoble). I got my PhD in October 2013, at the PERCEPTION team at INRIA Grenoble Rhône-Alpes, under the supervision of Prof. Radu Horaud, within the framework of the HUMAVIPS project. Nowadays I am a post-doctoral fellow in the Multimodal Human Understanding Group, in Trento, with Prof. Nicu Sebe.