{"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/deep-feature-learning-for-eeg-recordings","title":"Deep Feature Learning for EEG Recordings","arxiv_id":"1511.04306","date":"2015-11-13","proceeding":null,"authors":["Sebastian Stober","Avital Sternin","Adrian M. Owen","Jessica A. Grahn"],"abstract":"We introduce and compare several strategies for learning discriminative\nfeatures from electroencephalography (EEG) recordings using deep learning\ntechniques. EEG data are generally only available in small quantities, they are\nhigh-dimensional with a poor signal-to-noise ratio, and there is considerable\nvariability between individual subjects and recording sessions. Our proposed\ntechniques specifically address these challenges for feature learning.\nCross-trial encoding forces auto-encoders to focus on features that are stable\nacross trials. Similarity-constraint encoders learn features that allow to\ndistinguish between classes by demanding that two trials from the same class\nare more similar to each other than to trials from other classes. This\ntuple-based training approach is especially suitable for small datasets.\nHydra-nets allow for separate processing pathways adapting to subsets of a\ndataset and thus combine the advantages of individual feature learning (better\nadaptation of early, low-level processing) with group model training (better\ngeneralization of higher-level processing in deeper layers). This way, models\ncan, for instance, adapt to each subject individually to compensate for\ndifferences in spatial patterns due to anatomical differences or variance in\nelectrode positions. The different techniques are evaluated using the publicly\navailable OpenMIIR dataset of EEG recordings taken while participants listened\nto and imagined music.","url_abs":"http://arxiv.org/abs/1511.04306v4","url_pdf":"http://arxiv.org/pdf/1511.04306v4.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":"deep-feature-learning-for-eeg-recordings","repo_url":"https://github.com/brainhack-school2020/BHS-AuditoryMultimodal","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"CC0-1.0"}}],"tasks":[{"task_slug":"eeg-1","task_name":"EEG"},{"task_slug":"eeg","task_name":"Electroencephalogram (EEG)"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1511.04306","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}