SIGMA-PPG: Statistical-prior Informed Generative Masking Architecture for PPG Foundation Model

Zongheng Guo, Tao Chen, Yang Jiao, Yi Pan, Xiao Hu, Manuela Ferrario
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:37817-37849, 2026.

Abstract

Current foundation model for photoplethysmography (PPG) signals is challenged by the intrinsic redundancy and noise of the signal. Standard masked modeling often yields trivial solutions while contrastive methods lack morphological precision. To address these limitations, we propose a Statistical-prior Informed Generative Masking Architecture (SIGMA-PPG), a generative foundation model featuring a prior-guided adversarial masking mechanism, where a reinforcement learning-driven teacher leverages statistical priors to create challenging learning paths that prevent overfitting to noise. We also incorporate a semantic consistency constraint via vector quantization to ensure that physiologically identical waveforms—even those altered by recording artifacts or minor perturbations—map to shared indices. This enhances codebook semantic density and eliminates redundant feature structures. Pre-trained on over 120,000 hours of data, SIGMA-PPG achieves superior average performance compared to five state-of-the-art baselines across 12 diverse downstream tasks. The code and model weights are available at https://github.com/ZonghengGuo/SigmaPPG.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-guo26a, title = {{SIGMA}-{PPG}: Statistical-prior Informed Generative Masking Architecture for {PPG} Foundation Model}, author = {Guo, Zongheng and Chen, Tao and Jiao, Yang and Pan, Yi and Hu, Xiao and Ferrario, Manuela}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {37817--37849}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/guo26a/guo26a.pdf}, url = {https://proceedings.mlr.press/v306/guo26a.html}, abstract = {Current foundation model for photoplethysmography (PPG) signals is challenged by the intrinsic redundancy and noise of the signal. Standard masked modeling often yields trivial solutions while contrastive methods lack morphological precision. To address these limitations, we propose a Statistical-prior Informed Generative Masking Architecture (SIGMA-PPG), a generative foundation model featuring a prior-guided adversarial masking mechanism, where a reinforcement learning-driven teacher leverages statistical priors to create challenging learning paths that prevent overfitting to noise. We also incorporate a semantic consistency constraint via vector quantization to ensure that physiologically identical waveforms—even those altered by recording artifacts or minor perturbations—map to shared indices. This enhances codebook semantic density and eliminates redundant feature structures. Pre-trained on over 120,000 hours of data, SIGMA-PPG achieves superior average performance compared to five state-of-the-art baselines across 12 diverse downstream tasks. The code and model weights are available at https://github.com/ZonghengGuo/SigmaPPG.} }
Endnote
%0 Conference Paper %T SIGMA-PPG: Statistical-prior Informed Generative Masking Architecture for PPG Foundation Model %A Zongheng Guo %A Tao Chen %A Yang Jiao %A Yi Pan %A Xiao Hu %A Manuela Ferrario %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-guo26a %I PMLR %P 37817--37849 %U https://proceedings.mlr.press/v306/guo26a.html %V 306 %X Current foundation model for photoplethysmography (PPG) signals is challenged by the intrinsic redundancy and noise of the signal. Standard masked modeling often yields trivial solutions while contrastive methods lack morphological precision. To address these limitations, we propose a Statistical-prior Informed Generative Masking Architecture (SIGMA-PPG), a generative foundation model featuring a prior-guided adversarial masking mechanism, where a reinforcement learning-driven teacher leverages statistical priors to create challenging learning paths that prevent overfitting to noise. We also incorporate a semantic consistency constraint via vector quantization to ensure that physiologically identical waveforms—even those altered by recording artifacts or minor perturbations—map to shared indices. This enhances codebook semantic density and eliminates redundant feature structures. Pre-trained on over 120,000 hours of data, SIGMA-PPG achieves superior average performance compared to five state-of-the-art baselines across 12 diverse downstream tasks. The code and model weights are available at https://github.com/ZonghengGuo/SigmaPPG.
APA
Guo, Z., Chen, T., Jiao, Y., Pan, Y., Hu, X. & Ferrario, M.. (2026). SIGMA-PPG: Statistical-prior Informed Generative Masking Architecture for PPG Foundation Model. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:37817-37849 Available from https://proceedings.mlr.press/v306/guo26a.html.

Related Material