Established in 1998, the ALCOR Lab at the La Sapienza University of Rome conducts various research activities related to the perceptual inference processes that can be activated by cognitive robotics systems.
The ALCOR Lab aims at studying and developing the methods of representation as well as techniques of inference associated with the perceptual process. Using early attention as the “starting point” for research, the lab studies several aspects of audio-visual salience, which are typically induced by actions, motion as well as interactions. The lab is particularly interested in understanding how classifying the various aspects of the salience concept could help recognize several patterns of human behaviors while giving a better understanding of how events unfold in our surroundings.
The lab aims at achieving its goal by studying the different aspects of the entire recognition process, 3D reconstruction, interpretation of human facial expressions, motion interpretation, analysis of shapes, attention in audio-visual settings, and probabilistic as well as logical models of actions’ prediction. A majority of the experiments make use of the Gaze machine, a special device that is constantly developing.