eess.SPNov 4, 2026

Adapting Sleep Foundation Models for Dense Sleep Spindle Localization

Authors: Alexis Bollengier, Abel Díaz Berenguer, Olivier Pallanca, André Anjos, Hichem Sahli

Organizations: Department of Electronics and Informatics (ETRO), Vrije Universiteit Brussel (VUB), Brussels, Belgium · Idiap Research Institute, Martigny, Switzerland · Interuniversity Microelectronics Centre (IMEC), Heverlee, Belgium

Abstract

Adapting sleep foundation models to dense electroencephalography (EEG) event localization is challenging because multi-second representations may not preserve sufficient temporal detail for precise boundary recovery. We address sleep spindle localization by adapting a frozen SleepFM encoder with a local time-frequency pathway that preserves spindle-relevant spectral content and intra-token position. The local representation drives localization, while residual cross-attention incorporates SleepFM temporal context. Dense predictions are obtained through localized token-to-sample reconstruction and temporal refinement. Tested on the Massive Online Data Annotation (MODA), the proposed model achieves an event-level F1F_1 of 0.8080±0.00400.8080 \pm 0.0040 at IoU 0.2 across six folds. Ablations show that localization primarily depends on local spectral and sub-token temporal information, while the pretrained representation provides complementary context. These results indicate that dense adaptation benefits from preserving information below the pretrained token scale. Code will be released upon acceptance.

Figures & tables

Explore similar work

CardsList
  1. ZIPBrain: Can EEG Foundation Models Be Faster, Locally Deployable, but Accurate?

    Aug 7, 2026Lingwei Li, Yirong Kan, Peng Chen +3Efficient Transformer InferenceToken Merging

  2. Comparative Analysis of State-of-the-Art Foundation Models for Sleep Analysis Under Channel Reduction

    Aug 14, 2026Hassan Mehdi, Riku Klen, Ayse Kosal Bulbul +5Sleep Stage ClassificationEEG Foundation Models

  3. BrainCodes: A Lightweight EEG Foundation Model over Discrete Neural Codes

    Oct 21, 2026Kleanthis Avramidis, Ivan Tashev, Dimitra EmmanouilidouEEG Foundation ModelsRepresentation Learning