DANCE: Detect and Classify Events in EEG

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Lévy, Jarod, Banville, Hubert, Rapin, Jérémy, King, Jean-Remi, Moreau, Thomas, d'Ascoli, Stéphane
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866918494645256192
author Lévy, Jarod
Banville, Hubert
Rapin, Jérémy
King, Jean-Remi
Moreau, Thomas
d'Ascoli, Stéphane
author_facet Lévy, Jarod
Banville, Hubert
Rapin, Jérémy
King, Jean-Remi
Moreau, Thomas
d'Ascoli, Stéphane
contents Event identification in continuous neural recordings is a critical task in neuroscience. Decoding in EEG is dominated by classifying windows aligned to known event onsets. However, while available in controlled experiments, such onsets are absent in continuous real-world monitoring. Here, we introduce DANCE, a deep learning pipeline that frames neural decoding as a set-prediction problem and jointly detects and classifies events directly from raw, unaligned signals. Evaluated separately on ten datasets curated from the literature with a wide variety of event types (ranging from milliseconds to minutes in duration), our model outperforms existing methods on a broad range of cognitive, clinical and BCI tasks. This single architecture establishes a new state of the art in the competitive task of seizure monitoring and matches the accuracy of onset-informed models for BCI tasks. Overall, our method marks a step towards end-to-end asynchronous neural decoding models
format Preprint
id arxiv_https___arxiv_org_abs_2605_10688
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle DANCE: Detect and Classify Events in EEG
Lévy, Jarod
Banville, Hubert
Rapin, Jérémy
King, Jean-Remi
Moreau, Thomas
d'Ascoli, Stéphane
Machine Learning
Signal Processing
Event identification in continuous neural recordings is a critical task in neuroscience. Decoding in EEG is dominated by classifying windows aligned to known event onsets. However, while available in controlled experiments, such onsets are absent in continuous real-world monitoring. Here, we introduce DANCE, a deep learning pipeline that frames neural decoding as a set-prediction problem and jointly detects and classifies events directly from raw, unaligned signals. Evaluated separately on ten datasets curated from the literature with a wide variety of event types (ranging from milliseconds to minutes in duration), our model outperforms existing methods on a broad range of cognitive, clinical and BCI tasks. This single architecture establishes a new state of the art in the competitive task of seizure monitoring and matches the accuracy of onset-informed models for BCI tasks. Overall, our method marks a step towards end-to-end asynchronous neural decoding models
title DANCE: Detect and Classify Events in EEG
topic Machine Learning
Signal Processing
url https://arxiv.org/abs/2605.10688