Real-time estimation of eye gaze by in-ear electrodes.

Cognitive control of a hearing aid is the topic for several ongoing studies. The relevance of these studies should be seen in the light of inadequate steering of current hearing aids. While most studies are concerned with auditory attention tracking from the electroencephalogram (EEG), a complimentary approach may be to use visual attention tracking to steer the devices. Visual attention may be characterized by gaze direction, which can be obtained by electrooculography (EOG). EOG may be recorded from electrodes placed in the ear canal, termed EarEOG. To test the comparison of conventional EOG and EarEOG recordings, we conducted two experiments with six subjects. In the first experiment, the subjects were instructed to follow a moving dot on the screen moving in large saccades. In the second experiment, there were five large targets, and within each target, the dot had minor movements. When comparing conventional EOG and EarEOG, correlations of 0.9 and 0.91 with standard deviations of 0.02 were obtained for the two experiments respectively. To assess the feasibility of using EarEOG in real-time, correlation between EarEOG and the timecourse of the dot position was performed. When both signals were filtered with the same real-time applicable filter, correlations of 0.83 and 0.85 with standard deviations of 0.09 and 0.05 were found respectively to the two experiments. In conclusion, this study provides motivational aspects of using EarEOG to estimate eye gaze, as well as it identifies important future challenges in real-time applications to steer external devices such as a hearing aid.

A biologically inspired meta-control navigation system for the Psikharpax rat robot.

A biologically inspired navigation system for the mobile rat-like robot named Psikharpax is presented, allowing for self-localization and autonomous navigation in an initially unknown environment. The ability of parts of the model (e.g. the strategy selection mechanism) to reproduce rat behavioral data in various maze tasks has been validated before in simulations. But the capacity of the model to work on a real robot platform had not been tested. This paper presents our work on the implementation on the Psikharpax robot of two independent navigation strategies (a place-based planning strategy and a cue-guided taxon strategy) and a strategy selection meta-controller. We show how our robot can memorize which was the optimal strategy in each situation, by means of a reinforcement learning algorithm. Moreover, a context detector enables the controller to quickly adapt to changes in the environment-recognized as new contexts-and to restore previously acquired strategy preferences when a previously experienced context is recognized. This produces adaptivity closer to rat behavioral performance and constitutes a computational proposition of the role of the rat prefrontal cortex in strategy shifting. Moreover, such a brain-inspired meta-controller may provide an advancement for learning architectures in robotics.