Arvind Pillai

profile_pic.jpg

I am a Student Researcher at Google and a 4th year PhD student at Dartmouth College, advised by Prof. Andrew Campbell. Previously, I was a research intern at Nokia Bell Labs (Cambridge, UK) (Summer 2024), and an AI & Data Graduate Scientist at AstraZeneca (2019-2021).

My interdisciplinary research centers on developing machine learning techniques to study health in real-world settings, utilizing data from wearable devices and mobile phone sensors. In particular, I am interested in these areas:

  1. Health Foundation Models: Developing models that can generalize to many tasks is crucial for healthcare. To this end, I have developed PaPaGei, which is the first open-source foundation model for Photoplethysmography (PPG) [ICLR’25]; And proposed Time2Lang, which is a framework to integrate time-series foundation models with LLMs [CHIL’25].

  2. Un/Semi-supervised Learning: I developed a method to detect rare life events from multi-modal sensing data [CHIL’23]. Additionally, I have also proposed a domain adaptation method and investigated the generalization of speech-based suicidal ideation detection [UbiComp’24a].

  3. Behavior Modelling using Mobile Systems: I develop machine learning models and contribute to mobile systems targeted at understanding mental health. In this area, I have developed models for depression detection from in-the-wild smartphone images [CHI’24], an intervention system for improving social interactions in individuals with schizophrenia [IEEE PerCom Magazine’24], and maintain large-scale datasets with over 3 years of continuous sensing data [UbiComp’24b].

news

Jun, 2025 Excited to share that I’m starting my Student Researcher position at Google in Cambridge, MA! If you’re in the area, I’d love to grab coffee.
Apr, 2025 Delighted to share that Time2Lang is accepted to CHIL’25!
Jan, 2025 :parrot: PaPaGei is accepted to ICLR 2025!
Dec, 2024 :parrot: PaPaGei wins the :trophy: Best Paper Award at the Time Series in the Age of Large Models workshop at NeurIPS 2024.
Oct, 2024 :parrot: PaPaGei, a foundation model for PPG data is now available! Paper, Models, and Code
Oct, 2024 :microphone: Presented “Investigating Generalizability of Speech-Based Suicidal Ideation Detection Using Mobile Phones” @ Ubicomp 2024
Jun, 2024 :bell: I will be interning at Nokia Bell Labs in Cambridge, UK for Summer 2024
May, 2024 :microphone: Presented our work MoodCapture at CHI 2024 in Honolulu

Selected Publications

  1. CHIL
    Time2Lang: Bridging Time-Series Foundation Models and Large Language Models for Health Sensing Beyond Prompting
    Arvind Pillai, Dimitris Spathis, Subigya Nepal, and 6 more authors
    Conference on Health, Inference, and Learning, 2025
  2. ICLR
    PaPaGei: Open Foundation Models for Optical Physiological Signals
    Arvind Pillai, Dimitris Spathis, Fahim Kawsar, and 1 more author
    In International Conference on Learning Representations, 2025
  3. CHI
    MoodCapture: Depression Detection Using In-the-Wild Smartphone Images
    Subigya Nepal*, Arvind Pillai*, Weichen Wang, and 8 more authors
    In Proceedings of the CHI Conference on Human Factors in Computing Systems , 2024
  4. UbiComp
    Investigating Generalizability of Speech-based Suicidal Ideation Detection Using Mobile Phones
    Arvind Pillai, Subigya Kumar Nepal, Weichen Wang, and 8 more authors
    Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies, 2024