Recognition and Analysis of Colours Using Generative AI And Brain Computer Interface

Identification of colour is a significant cognitive ability, especially in human beings. Analysing this ability through Brain-Computer Interface (BCI) technology offers insights into the neurocognitive mechanisms underlying colour perception. In this work, the effect of colour in brain’s response to a visual stimulus is explored. The three primary colours- red, green, blue and also black were applied as visual stimuli to 5 healthy participants, the EEG signals were recorded using Ultracortex Mark 4 and a novel dataset was created. The signals were analysed using a 5layered Keras sequential deep neural network and the performance was evaluated. The classifier performance of BCI system depends greatly on the quantity as well as quality of the recorded EEG data. The distraction of the users’ attention therefore affects the output and to improve classifier robustness, a solution is the generation of artificial data. In this work, generated EEG is used to augment the training data for effective classification. We hope that the outcomes of this study will throw light to the neurological background of colour perception.