Dimitris Spathis
Dimitris Spathis is a Research Scientist in the Google Health Research team and a Visiting Academic at the University of Cambridge. His research sits at the intersection of multimodal foundation models and AI for health, focusing on building data-efficient and robust machine learning systems that can handle the complexities of real-world personal data. At Google, he develops physiological sensing models for wearable devices, such as the Pixel Watch and Fitbit, and contributes to the scaling of foundation models for continuous health monitoring. Prior to Google, Dimitris was a Senior Research Scientist at Nokia Bell Labs, where he led the release of PaPaGei, the first open foundation model for biosignals. He holds a PhD in AI from the University of Cambridge. His research is regularly published in leading venues across AI and human-centered signal processing—including NeurIPS, ICLR, KDD, and Nature Digital Medicine—and has been covered by major media outlets including The New York Times, BBC, and Forbes. Website: https://dispathis.com/
Authored Publications
Sort By
Pixel Watch: Robust Heart Rate Sensing from Multipath PPG and On-Device Deep Learning Trained on 10,000 hours of Free-Living and Fitness Data
Megan Walker
Yojan Patel
Shyam Tailor
Matt Wimmer
Brennan Garrett
Dan Howe
Abhinuv Pitale
Hamed Vavadi
Tien Le
Steve Diamond
Oleksiy Vyalov
Vik Sharma
Pete Richards
Tracy Giest
Erika Siegel
Tuan Phan
Sam Mravca
Derrick Vickers
Benjamin Stone
Katarina Vukosavljević
Justin Phillips
YongSuk Cho
Stefanie Hollidge
Antony Siahaan
Soren Brage
Shwetak Patel
Robert Harle
IEEE Sensors Letters (2026)
Preview abstract
The Pixel Watch 2 (PW2) is the first Google smartwatch to combine multipath photoplethysmography (PPG) with deep learning-based heart rate inference, designed to significantly improve sensing accuracy during motion-heavy activities. The device processes 10 optical channels using an on-device, 15-layer temporally dilated convolutional neural network (~300K parameters) to yield a 1 Hz heart rate output. Crucial to this model's performance was its training on a massive dataset comprising 10,000 hours of data from 962 participants, curated from a broader corpus of controlled and free-living activities. We evaluated the PW2's sensing performance across two independent validation sets: an in-house fitness dataset (229 participants, 250 hours) and an external free-living dataset (27 participants, 1000+ hours). The system achieved 95% Limits of Agreement of -10.34 to 8.66 BPM during exercise and -6.57 to 7.48 BPM during free-living activities, demonstrating substantially tighter error margins than previous Google devices. Finally, we discuss key design lessons, emphasizing that large-scale deep learning was instrumental in fully leveraging multipath PPG hardware over traditional signal processing approaches.
View details
