Retrieval from out-of-domain foundation models enables personalization of a lightweight transformer for stress detection, yielding +3.92% accuracy and +4.76% F1 gains on WESAD without user labels.
arXiv preprint arXiv:2504.19596 , year=
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
years
2026 4verdicts
UNVERDICTED 4representative citing papers
PAMF initializes flow matching with missingness-type priors and shares encoder weights between imputation and classification to improve multimodal time-series prediction under incomplete observations.
A compact 0.09B model using hierarchical discrete tokenization and prompted latent translation outperforms larger baselines in cross-modal PPG-to-ECG synthesis and cross-frequency super-resolution.
AMI reduces sensor usage by 48.8% and improves accuracy by 1.9% on average across three medical datasets by jointly learning when to sense and how to infer from multimodal physiological signals.
citing papers explorer
-
Retrieval-Augmented Personalization with Foundation Models for Wearable Stress Detection
Retrieval from out-of-domain foundation models enables personalization of a lightweight transformer for stress detection, yielding +3.92% accuracy and +4.76% F1 gains on WESAD without user labels.
-
PAMF: Prior-Aware Multimodal Fusion for Incomplete Time Series Data
PAMF initializes flow matching with missingness-type priors and shares encoder weights between imputation and classification to improve multimodal time-series prediction under incomplete observations.
-
Compact Latent Manifold Translation: A Parameter-Efficient Foundation Model for Cross-Modal and Cross-Frequency Physiological Signal Synthesis
A compact 0.09B model using hierarchical discrete tokenization and prompted latent translation outperforms larger baselines in cross-modal PPG-to-ECG synthesis and cross-frequency super-resolution.
-
Sense Less, Infer More: Agentic Multimodal Transformers for Edge Medical Intelligence
AMI reduces sensor usage by 48.8% and improves accuracy by 1.9% on average across three medical datasets by jointly learning when to sense and how to infer from multimodal physiological signals.