{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/compact-convolutional-neural-networks-for","title":"Compact Convolutional Neural Networks for Classification of Asynchronous Steady-state Visual Evoked Potentials","arxiv_id":"1803.04566","date":"2018-03-12","proceeding":null,"authors":["Nicholas R. Waytowich","Vernon Lawhern","Javier O. Garcia","Jennifer Cummings","Josef Faller","Paul Sajda","Jean M. Vettel"],"abstract":"Steady-State Visual Evoked Potentials (SSVEPs) are neural oscillations from\nthe parietal and occipital regions of the brain that are evoked from flickering\nvisual stimuli. SSVEPs are robust signals measurable in the\nelectroencephalogram (EEG) and are commonly used in brain-computer interfaces\n(BCIs). However, methods for high-accuracy decoding of SSVEPs usually require\nhand-crafted approaches that leverage domain-specific knowledge of the stimulus\nsignals, such as specific temporal frequencies in the visual stimuli and their\nrelative spatial arrangement. When this knowledge is unavailable, such as when\nSSVEP signals are acquired asynchronously, such approaches tend to fail. In\nthis paper, we show how a compact convolutional neural network (Compact-CNN),\nwhich only requires raw EEG signals for automatic feature extraction, can be\nused to decode signals from a 12-class SSVEP dataset without the need for any\ndomain-specific knowledge or calibration data. We report across subject mean\naccuracy of approximately 80% (chance being 8.3%) and show this is\nsubstantially better than current state-of-the-art hand-crafted approaches\nusing canonical correlation analysis (CCA) and Combined-CCA. Furthermore, we\nanalyze our Compact-CNN to examine the underlying feature representation,\ndiscovering that the deep learner extracts additional phase and amplitude\nrelated features associated with the structure of the dataset. We discuss how\nour Compact-CNN shows promise for BCI applications that allow users to freely\ngaze/attend to any stimulus at any time (e.g., asynchronous BCI) as well as\nprovides a method for analyzing SSVEP signals in a way that might augment our\nunderstanding about the basic processing in the visual cortex.","url_abs":"http://arxiv.org/abs/1803.04566v2","url_pdf":"http://arxiv.org/pdf/1803.04566v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"compact-convolutional-neural-networks-for","repo_url":"https://github.com/vlawhern/arl-eegmodels","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"compact-convolutional-neural-networks-for","repo_url":"https://github.com/jinglescode/python-signal-processing","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"eeg-1","task_name":"EEG"},{"task_slug":"eeg","task_name":"Electroencephalogram (EEG)"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"ssvep","task_name":"SSVEP"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1803.04566","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}