When we inspect a scene the visual system processes some aspects of the scene in more detail than others. In order to use the brains limited capacities in an optimal manner a big proportion of the information available on the retina needs to be filtered out by the mechanism of visual attention. Attention therefore means facilitated visual processing of some information and discarding of other information. How does the visual system weight its inputs?
The way attention is deployed in a scene crucially depends on the task at hand. Arguably the most important function of the visual system is to guide our behavior, and our movements in particular. A special research focus lies therefore on the planning and preparation of movements and on how our movement intentions influence the way we perceive. In visually guiding our movements the brain acts as a sensory-motor interface that transforms the visual information about a movement goal into specific motor commands. In order to do this most efficiently the attentional system preferentially processes information that is relevant and important for the planned movement. This becomes especially important in natural environments that are full of task-irrelevant objects and distractors.
This task-dependent weighting of visual input becomes an especially fascinating research topic when investigating more complexmovement tasks. Many movements we perform everyday require to consider more than just a single motor goal. For example, we often have to avoid obstacles or we want to combine several movement goals within one common action plan. We often concatenate movements and prepare for movement sequences. Our recent investigations showed that the attentional system flexibly handles the requirements that are posed by those tasks: the focus of visual attention can be split among several objects of importance, forming attentional landscapes within the visual scene.
From a physiological perspective, planning areas in frontal and parietal planning areas play a crucial role in weighting incoming visual signals by providing top-down feedback signals to areas involved in visual processing. Recently we showed, for example, that individual cells within the posterior parietal cortex (PPC) encode in parallel multiple subsequent movement goals for a planned reach sequence.