Flickering source is an indispensable component in steady-state visual evoked potentials (SSVEPs)-based brain-computer interface (BCI), and its background severely influences the potentials evoked by the repetitive stimuli. In this paper, we investigated the problem under three different backgrounds in the context of the SSVEP-BCI-based robot car control, including black screen, static scene and dynamic scene of the environment. In the ten subjects experiment, we found significant decrease in SSVEP amplitude in dynamic scene condition compared to the reference condition black screen (p < 0.05), which resulted in classification accuracy decrease as evaluated by 10-fold cross validation. However, our proposed experiment paradigm has shown that training with static scene or dynamic scene condition could well compensate this performance drop and improve the online robot car control with real-time video feedback. The addressed problem in our application would provide some valuable suggestions when translating the SSVEP-BCI from laboratory exploration into practical usages.
Bibliographical noteFunding Information:
We thank all volunteers for their participation in the study. This paper is supported by the National Basic Research Program (973 Program) of China (Grant No. 2011CB013305), the Science and Technology Commission of Shanghai Municipality (Grant No. 13430721600), and the National Natural Science Foundation of China (Grant Nos. 51375296, 51421092).
- Stimulation background
- robot car control