Research Group: Cybernetics Approach to Perception and Action research (CAPA)
Supervisors: Jean-Pierre Bresciani, Lewis Chuang

I study human eye movements to examine where and how our eyes move to access task-relevant information (e.g., while steering a vehicle or during visual search). In this context, I've been working in the BW-FIT project information at your fingertips and in the myCopter project.

The effects of perceptual decisions on saccadic eye movements

Introduction

Existing models of gaze control primarily include basic aspects of the visual scene or the biomechanics of the eyes. What is currently missing is a better consideration of top-down factors such as task knowledge or motivation. Such models are useful in a variety of areas, for instance, in the design and evaluation of human-machine interfaces.

Goals

We investigate how human eye movements are shaped by task-properties, more specifically, the properties of the perceptual decisions which they support. This includes, for instance, the expected difficulty or the consequences of a decision in terms of the reward or cost of ensuing actions. This will lay the groundwork for more advanced models of human gaze control.

Methods

We employ high-speed oculometry to measure saccade latency, kinematics, and fixations in object recognition and dual-task paradigms. For instance, (i) a standard saccade paradigm is combined with a 2-alternative forced-choice recognition task and (ii) a saccade and recognition task is carried out in parallel to a compensatory tracking task.

Initial results

Previous work postulated task-specific influences on saccade characteristics (e.g., due to the need for eye and hand coordination). We tested whether similar effects occur in saccades that support a basic visual task. Our results show changes in saccade characteristics: Task-related saccades are initiated earlier and are faster than classical target-elicited saccades.

Initial conclusion

These results suggest a basic unifying explanation for some of the previously measured saccade changes. Existing research on the reward-sensitivity of the oculomotor system leads us to the hypothesis that a closer analysis of a task's reinforcement-structure could potentially explain such task-related effects. Experiments pursuing this idea are currently ongoing.

I studied computer science at the University of Konstanz (M.Sc.), where I specialized on human-computer interaction. In Harald Reiterer's group I worked on new interaction techniques and experimental pointing device performance tests (along the lines of ISO 9241-9). This work was partially funded by the information technology program by the state of Baden-Württemberg (BW-Fit).

In 2009 I joined Heinrich Bülthoff's group at the MPI for Biological Cybernetics as a Ph.D. student to study eye movements during interactive tasks. Here, I started to work with Lewis Chuang and Jean-Pierre Bresciani (CNRS Grenoble) on gaze behavior during object recognition and motor control.