A Novel Steady-State Visually Evoked Potential-Based Brain-Computer-Interface Paradigm to Steer a Humanoid Robot

Abstract

In this paper, a brain-computer interface (BCI) system based on steady-state visually evoked EEG potentials (SSVEP) has been presented to steer a NAO humanoid robot, in which a novel robot control paradigm with two-layer-interface was designed and implemented. The scene captured from robot camera was divided into different target regions and displayed to the human subjects with extra visual stimulus superposed on it to reduce the influence of dynamic background scene and the side-effect of gaze shift. The visually evoked potentials was captured from the subjects' EEG signals to accomplish the directions selection mission and to control the NAO robot's movements towards the target region. All the subjects could complete the mission using our BCI paradigm in the real-world environment. The results showed the feasibility of our BCI paradigm in brain-actuated robot control.