A closed-loop EEG-based visual stimulation framework from controllable generation
Keywords: Neural modulation; EEG; Close-loop;
Abstract: Recent advancements in artificial neural networks (ANNs) have significantly refined methodologies for predicting the neural coding activities of the ventral visual stream in human and animal brains based on visual stimuli. Nevertheless, the endeavor to control visual stimuli to elicit specific neural activities continues to confront substantial challenges, including prohibitive experimental costs, the high-dimensional nature of stimuli, pronounced inter-individual variability, and an incomplete understanding of neuronal selectivity. To address these impediments, we propose a novel electroencephalography (EEG)-based closed-loop framework for visual stimulus. Leveraging this framework, we can identify the optimal natural image stimulus within a theoretically infinite search space to maximize the elicitation of neural activities that most closely align with desired brain states. Our framework employs advanced ANN ensemble models to ensure the reliability of neural activity predictions. Furthermore, we conceptualize the brain coding predicted by the ANN model as a non-differentiable black-box process, allowing us to directly analyze the relationship between the administered visual stimuli and the targeted brain activity. Our research demonstrates that, independent of the exactness of the ANN-predicted brain coding, the proposed framework can procure the theoretically optimal natural image stimulus at given cycle steps. Moreover, our method exhibits generalizability across different modalities of brain-specific activity regulation. Our code is available at https://anonymous.4open.science/status/closed-loop-F2E9.
Primary Area: applications to neuroscience & cognitive science
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.
Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 134
Loading