The Distinction Between Object Recognition and Object Identification in Brain Connectivity for Brain–Computer Interface Applications

Object recognition and object identification are complex cognitive processes where information is integrated and processed by an extensive network of brain areas. However, although object recognition and object identification are similar, they are considered separate functions in the brain. Interestingly, the difference between object recognition and object identification has still not been characterized in a way that brain–computer interface (BCI) applications can detect or use. Hence, in this study, we investigated neural features during object recognition and identification tasks through functional brain connectivity. We conducted an experiment involving 25 participants to explore these neural features. Participants completed two tasks: an object recognition task, where they determined whether a target object belonged to a specified category, and an object identification task, where they identified the target object among four displayed images. Our aim was to discover reliable features that could distinguish between object recognition and identification. The results demonstrate a significant difference between object recognition and identification in the participation coefficient (PC) and clustering coefficient (CC) of delta activity in the visual and temporal regions of the brain. Further analysis at the category level shows that this coefficient differs for different categories of objects. Utilizing these discovered features for binary classification, the accuracy for the animal category reached 80.28%. The accuracy for flower and vehicle categories also improved when combining the PC and CC, although no improvement was observed for the food category. Overall, what we have found is a feature that might be able to be used to differentiate between object recognition and identification within a BCI object recognition system. Further, it may help BCI object recognition systems to determine a user’s intentions when selecting an object.