Personal tools
You are here: Home Publications Visual Saliency in the Wild: Moving the Analysis of Gaze Behaviors to Three-Dimensional, Unstructured Environments
About NIFTi
NIFTi is about human-robot cooperation. About teams of robots and humans doing tasks together, interacting together to try and reach a shared goal. NIFTi looks at how the robot could bear the human in mind. Literally. When determining what to do or say next in human-robot interaction; when, and how. NIFTi puts the human factor into cognitive robots, and human-robot team interaction in particular.   Each year, NIFTi evaluates its systems together with several USAR organizations. Rescue personnel teams up with NIFTi robots to carry out realistic missions, in real-life training areas. 
Impressum

This site uses Google Analytics to record statistics on site visits - see Legal information.

 

M. Pizzoli (2011)

Visual Saliency in the Wild: Moving the Analysis of Gaze Behaviors to Three-Dimensional, Unstructured Environments

PhD thesis, Department of Computer Science, La Sapienza University.

The study of Visual Attention has emerged as a new, inter-disciplinary research area, crucial for understanding vision. A fundamental role is played by eye tracking experiments, which have been extensively used to study human selection mechanisms and promoted the development of computational models of visual attention, whose well known outcomes are the saliency maps. Among the eye trackers, wearable ones have the advantages of allowing the estimation of the Point of Regard (POR) while performing natural tasks, instead of experimental, static lab settings. Indeed, if, on one hand, eye tracking experiments have been performed extensively to study visual attention, on the other hand it is becoming clear that those stimuli often used in laboratory experiments, such as still images or videos that have been suitably edited, may not be representative of natural viewing behaviors. The observed scenes usually contains dynamic, three-dimensional stimuli. Moreover, the observer is free to move and data association requires analysis in 3D in order to take into account that, in presence of ego motion, an object of interest can be attended from different points of view. Thus, moving the analysis of gaze data from static, eye tracking experiments to 3D, dynamic and real scenarios requires a new framework for the identification, localization and tagging of human salient gaze behaviors. The data acquisition relies on a wearable device, which comprises four cameras: two cameras acquire the eye movements while a pair of cameras in stereo configuration acquire the scene.
Document Actions