FedHUG: Federated Heterogeneous Unsupervised Generalization for Remote Physiological Measurements

2025/10/14 by Xiao Yang, Yang, Xiao, He, Dengbo +2
Engineering · Medicine · #Computer Vision and Pattern Recognition (cs.CV) #ECG Monitoring and Analysis #FOS: Computer and information sciences #Non-Invasive Vital Sign Monitoring

paper · pdf · doi:10.48550/arxiv.2510.12132

openalex publication_date 2025/10/14 · openalex created_date 2025/10/17 · openalex updated_date 2026/07/28

Abstract

Remote physiological measurement gained wide attention, while it requires collecting users' privacy-sensitive information, and existing contactless measurements still rely on labeled client data. This presents challenges when we want to further update real-world deployed models with numerous user data lacking labels. To resolve these challenges, we instantiate a new protocol called Federated Unsupervised Domain Generalization (FUDG) in this work. Subsequently, the Federated Heterogeneous Unsupervised Generalization (FedHUG) framework is proposed and consists of: (1) Minimal Bias Aggregation module dynamically adjusts aggregation weights based on prior-driven bias evaluation to cope with heterogeneous non-IID features from multiple domains. (2) The Global Distribution-aware Learning Controller parameterizes the label distribution and dynamically manipulates client-specific training strategies, thereby mitigating the server-client label distribution skew and long-tail issue. The proposal shows superior performance across state-of-the-art techniques in estimation with either RGB video or mmWave radar. The code will be released.

Citations

• FeatureFool: Zero-Query Fooling of Video Models via Feature Map
• Towards Better Dental AI: A Multimodal Benchmark and Instruction Dataset for Panoramic X-ray Analysis
• PhysDrive: A Multimodal Remote Physiological Measurement Dataset for In-vehicle Driver Monitoring
• Not Only Consistency: Enhance Test-Time Adaptation with Spatio-temporal Inconsistency for Remote Physiological Measurement
• Align the GAP: Prior-based Unified Multi-Task Remote Physiological Measurement Framework For Domain Generalization and Personalization
• Towards Generalizable Drowsiness Monitoring with Physiological Sensors: A Preliminary Study
• Geometric Knowledge-Guided Localized Global Distribution Alignment for Federated Learning
• CardiacMamba: A Multimodal RGB-RF Fusion Framework with State Space Models for Remote Physiological Measurement
• Revisiting Interactions of Multiple Driver States in Heterogenous Population and Cognitive Tasks
• Efficient Mixture-of-Expert for Video-based Driver State and Physiological Multi-task Estimation in Conditional Autonomous Driving
• Query-Efficient Video Adversarial Attack with Stylized Logo
• Fair Federated Learning under Domain Skew with Local Consistency and Domain Diversity
• Federated Unsupervised Domain Generalization using Global and Local Alignment of Gradients
• PhysMLE: Generalizable and Priors-Inclusive Multi-task Remote Physiological Measurement
• Rethinking the Representation in Federated Unsupervised Learning with Non-IID Data
• Advancing Generalizable Remote Physiological Measurement through the Integration of Explicit and Implicit Prior Knowledge
• GPT as Psychologist? Preliminary Evaluations for GPT-4V on Visual Affective Computing
• Probabilistic Contrastive Learning for Long-Tailed Visual Recognition
• SimPro: A Simple Probabilistic Framework Towards Realistic Long-Tailed Semi-Supervised Learning
• Heterogeneous Federated Learning: State-of-the-art and Research Challenges
• Non-Contrastive Unsupervised Learning of Physiological Signals from Video
• MMPD: Multi-Domain Mobile Video Physiology Dataset
• PhysFormer++: Facial Video-based Physiological Measurement with SlowFast Temporal Difference Transformer
• Federated Learning with Label Distribution Skew via Logits Calibration
• SCAMPS: Synthetics for Camera Measurement of Physiological Signals
• Divergence-aware Federated Self-Supervised Learning
• Federated Remote Physiological Measurement with Imperfect Data
• Federated Learning with Buffered Asynchronous Aggregation
• No Fear of Heterogeneity: Classifier Calibration for Federated Learning with Non-IID Data
• Federated Unsupervised Representation Learning
• Video-based Remote Physiological Measurement via Cross-verified Feature Disentangling
• Bootstrap your own latent: A new approach to self-supervised Learning
• PulseGAN: Learning to generate realistic pulse waveforms in remote photoplethysmography
• The future of digital health with federated learning
• Momentum Contrast for Unsupervised Visual Representation Learning
• Momentum Contrast for Unsupervised Visual Representation Learning
• Remote Heart Rate Measurement from Highly Compressed Facial Videos: an End-to-end Deep Learning Solution with Video Enhancement
• Federated Optimization in Heterogeneous Networks
• VIPL-HR: A Multi-modal Database for Pulse Estimation from Less-constrained Face Video
• DeepPhys: Video-Based Physiological Measurement Using Convolutional Attention Networks
• Algorithmic Principles of Remote PPG
• RhythmFormer: Extracting Patterned rPPG Signals based on Periodic Sparse Attention

Cited by

• PhysDrive: A Multimodal Remote Physiological Measurement Dataset for In-vehicle Driver Monitoring

Related

• Federated Remote Physiological Measurement with Imperfect Data
• Align the GAP: Prior-based Unified Multi-Task Remote Physiological Measurement Framework For Domain Generalization and Personalization
• HtFLlib: A Comprehensive Heterogeneous Federated Learning Library and Benchmark
• FedMFS: Federated Multimodal Fusion Learning with Selective Modality Communication
• Federated Learning for Time-Series Healthcare Sensing with Incomplete Modalities