Welcome to Human Vision Modelling's home page
Enhanced Media Program: IMX-WP6
Rennes Compression Lab
Research Overview:
Modeling of the bottom-up human visual attention:
The visual environment presents far more perceptual information than can be effectively processed. In order to keep the essential visual information, human beings have developed a particular strategy, firstly outlined by W. James, the father of the psychology in 1890. This strategy, confirmed during the last two decades, requires two kinds of mechanisms. The first mechanism refers to the sensory attention driven by environmental events, commonly called bottom-up or stimulus-driven. The second one is the voluntational attention to both external and internal stimuli, commonly called top-down or goal-driven. Developing computational models attempting to simulate the human visual attention is a great challenge since the potential applications are numerous. Among them, content retargeting, image and video coding, or human-machine interfaces are undeniably important targets.
As a result, designing computational models of the human visual attention has become an active research area in recent years. A first trend was to incorporate some well-understood Human Visual System (HVS) features in order to detect automatically the human region of interest.
The second research trend is to develop models of visual attention by integrating advanced knowledge about the HVS and by simulating the bottom-up mechanism involved in the HVS.
Particular examples coming from three artists:
Déjeuner sur l'herbe, Mannet
La Main, S. Dali
