Saved in:
Bibliographic Details
Main Authors: Weng, Yuxuan, Luo, Wenhan, Shao, Qijia
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.18837
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917509607718912
author Weng, Yuxuan
Luo, Wenhan
Shao, Qijia
author_facet Weng, Yuxuan
Luo, Wenhan
Shao, Qijia
contents Wearable devices enable continuous health monitoring from multimodal signals, but real-world deployment is hindered by limited labeled data and pervasive sensor incompleteness. While large-scale self-supervised pretraining reduces label dependence, most existing methods assume full modality availability. Current approaches for handling modality missingness often reconstruct entire absent signals, which can encourage hallucinating modality-specific details that are not inferable from the observed sensor signals and degrade robustness. We propose VCR, a self-supervised framework that learns to extract valid representations robust to modality missingness. VCR employs an orthogonal tokenizer to enforce strict orthogonal disentanglement by rectifying latent manifolds and applying a geometric projection, separating each modality into shared semantics and modality-specific residuals. This design preserves complete information integrity while serving as a structural foundation for robust learning under modality missingness. The resulting tokens are processed by a missing-aware mixture-of-experts backbone that adapts to varying patterns of modality availability. By constraining the objective to reconstruct only the shared components of missing modalities, VCR effectively mitigates hallucinations of non-inferable modality-specific details. Across multiple health monitoring tasks, VCR consistently improves performance and robustness under full, single-missing, and multiple-missing modality settings compared with strong supervised and self-supervised baselines.
format Preprint
id arxiv_https___arxiv_org_abs_2605_18837
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle VCR: Learning Valid Contextual Representation for Incomplete Wearable Signals
Weng, Yuxuan
Luo, Wenhan
Shao, Qijia
Machine Learning
Artificial Intelligence
Signal Processing
Wearable devices enable continuous health monitoring from multimodal signals, but real-world deployment is hindered by limited labeled data and pervasive sensor incompleteness. While large-scale self-supervised pretraining reduces label dependence, most existing methods assume full modality availability. Current approaches for handling modality missingness often reconstruct entire absent signals, which can encourage hallucinating modality-specific details that are not inferable from the observed sensor signals and degrade robustness. We propose VCR, a self-supervised framework that learns to extract valid representations robust to modality missingness. VCR employs an orthogonal tokenizer to enforce strict orthogonal disentanglement by rectifying latent manifolds and applying a geometric projection, separating each modality into shared semantics and modality-specific residuals. This design preserves complete information integrity while serving as a structural foundation for robust learning under modality missingness. The resulting tokens are processed by a missing-aware mixture-of-experts backbone that adapts to varying patterns of modality availability. By constraining the objective to reconstruct only the shared components of missing modalities, VCR effectively mitigates hallucinations of non-inferable modality-specific details. Across multiple health monitoring tasks, VCR consistently improves performance and robustness under full, single-missing, and multiple-missing modality settings compared with strong supervised and self-supervised baselines.
title VCR: Learning Valid Contextual Representation for Incomplete Wearable Signals
topic Machine Learning
Artificial Intelligence
Signal Processing
url https://arxiv.org/abs/2605.18837