{"id":"arxiv:2212.04951","kind":"preprint","title":"EEG-NeXt: A Modernized ConvNet for The Classification of Cognitive Activity from EEG","authors":["Andac Demir","Iya Khalil","Bulent Kiziltan"],"author_count":3,"year":2022,"published":"2022-12-08","venue":null,"cited_by":null,"doi":null,"arxiv_id":"2212.04951","nct_id":null,"source_url":"https://arxiv.org/abs/2212.04951","archive_url":"https://space-kitty.com/papers/arxiv:2212.04951","updated":"2022-12-08","pmid":null,"pmcid":null,"openalex_id":null,"pdf_url":"https://arxiv.org/pdf/2212.04951","abstract":"One of the main challenges in electroencephalogram (EEG) based brain-computer interface (BCI) systems is learning the subject/session invariant features to classify cognitive activities within an end-to-end discriminative setting. We propose a novel end-to-end machine learning pipeline, EEG-NeXt, which facilitates transfer learning by: i) aligning the EEG trials from different subjects in the Euclidean-space, ii) tailoring the techniques of deep learning for the scalograms of EEG signals to capture better frequency localization for low-frequency, longer-duration events, and iii) utilizing pretrained ConvNeXt (a modernized ResNet architecture which supersedes state-of-the-art (SOTA) image classification models) as the backbone network via adaptive finetuning. On publicly available datasets (Physionet Sleep Cassette and BNCI2014001) we benchmark our method against SOTA via cross-subject validation and demonstrate improved accuracy in cognitive activity classification along with better generalizability across cohorts.","abstract_license":"CC0-1.0 (arXiv descriptive metadata)","license":null,"is_open_access":false,"retracted":false,"peer_reviewed":false,"categories":["eess.SP","cs.HC","cs.LG","q-bio.NC"],"details":null,"provenance":"arxiv","attribution":"Abstract: arXiv descriptive metadata (CC0). Thank you to arXiv for use of its open access interoperability. This service was not reviewed or approved by, nor does it necessarily express or reflect the policies or opinions of, arXiv.","topics":[],"cited_in_dossiers":[]}