Cross-sensor observation and fused measurement systems.
Multimodal Sensing is the topic program focused on how multiple observation channels can be aligned, fused, and made robust to missingness and heterogeneity. It sits inside the Perception pillar as the research program for measurement beyond a single waveform or sensor stream.
Fusion
Cross-sensor alignment, synchronized observations, and the structure of shared latent measurement.
Robustness
Missing modalities, sensor failure, partial availability, and acquisition mismatch.
Foundation Models
Large-scale multimodal pretraining for physiological and sensing systems.
Multimodal Biosignal Foundation Models
Cross-modal pretraining and modality-robust physiological inference.
Self-Supervised Representation Learning for Human Physiological Data
Self-supervised objectives for ECG, EEG, and physiological data.
Signals
Signal-level foundations for multimodal measurement work.