Research map / Vision-language and multimodal / Cross-modal learning
sleep2vec: Unified Cross-Modal Alignment for Heterogeneous Nocturnal Biosignals
Weixuan Yuan, Zengrui Jin, Yichen Wang, Donglin Xie, Ziyi Ye, Chao Zhang, Xuesong Chen
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use sleep2vec instead of SleepFM foundation model | sleep2vec · SHHS ECG & ABD | accuracy (%): 77.9 -> 82.7 | Improved |
| foundation model vs specialized EEG-only models | sleep2vec · SHHS EEG only | accuracy (%): 87.7 -> 87.4 | No clear effect |
| add metadata-aware weighting (DASH-InfoNCE) instead of vanilla InfoNCE (ablation) | sleep2vec · APPLES unseen cohort, full channels | accuracy (%): 76.8 -> 78.4 | Improved |
| DASH-InfoNCE instead of vanilla InfoNCE (ablation) | sleep2vec · SHHS full channels | accuracy (%): 88.4 -> 88.6 | Improved |
| DASH-InfoNCE instead of vanilla InfoNCE (ablation) | sleep2vec · pre-training retrieval | retrieval recall@1: 0.351 -> 0.368 | Improved |
| gating fusion instead of mean fusion (ablation) | sleep2vec · APPLES unseen cohort | accuracy (%): 78.1 -> 78.4 | Improved |
| gating fusion instead of concatenation (ablation) | sleep2vec · APPLES unseen cohort | accuracy (%): 76.9 -> 78.4 | Improved |
| add 15% masking during pre-training (ablation) | sleep2vec · APPLES unseen cohort | macro-F1 (%): 64.0 -> 65.2 | Improved |
| raise masking ratio from 15% to 30% (ablation) | sleep2vec · APPLES unseen cohort | macro-F1 (%): 65.2 -> 63.4 | Hurt |
| increase pre-training data from 25% to 100% (ablation) | sleep2vec · SHHS sleep staging | accuracy (%): 87.1 -> 88.6 | Improved |
| scale model from 63.5M to 238.2M parameters (ablation) | sleep2vec · SHHS and WSC sleep staging | accuracy | Improved |
| increase number of input modalities (ablation) | sleep2vec · SHHS disease prediction | ROC-AUC | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- SleepFM — SleepFM: Multi-modal Representation Learning for Sleep Across Brain Activity, ECG and Respiratory Signals (ICML 2024) · better on 7 of 7 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- SleepFM: Multi-modal Representation Learning for Sleep Across Brain Activity, ECG and Respiratory Signals (ICML 2024 · similarity 0.88)
- Robust Sleep Staging over Incomplete Multimodal Physiological Signals via Contrastive Imagination (NeurIPS 2024 · similarity 0.88)
- Generalizable Sleep Staging via Multi-Level Domain Alignment (AAAI 2024 · similarity 0.87)
- Subject-Relative Micro-Motion and Sleep Dynamics for Near-Infrared Video Sleep Staging (NeurIPS 2026 · similarity 0.85)
- Personalized Sleep Staging Leveraging Source-free Unsupervised Domain Adaptation (AAAI 2025 · similarity 0.84)
- NeuroAtlas: Benchmarking Foundation Models for Clinical EEG and Brain-Computer Interfaces (NeurIPS 2026 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
