Search for: All records

The quality of parent-child interactions at an early age has been linked to children's social-emotional development, executive function, and risk for behavior problems. As such, parent-child interactions in naturalistic settings could present a unique opportunity to screen for at-risk behavior in young children, enabling timely and targeted interventions. In this work, we validate the feasibility of using structured at-home play sessions, completed via the Tandem smartphone app, to enable highly accurate and scalable behavioral assessments. We demonstrate that audio and physiological signals recorded during the play session can be used to capture key markers of parent-child interaction dynamics, which are more indicative of at-risk behavior compared to features from each individual alone. We propose novel audio-based dyadic interaction features that significantly outperform conventional speech features at predicting risk for behavior problems, achieving an F1 score of 0.87. Furthermore, we show that dyadic physiological synchrony features, extracted from privacy-preserving wearable sensor data, can classify at-risk behavior with an F1 score of 0.91. Tandem thus sets the stage for automated at-home behavior assessment tools for young children that balance screening accuracy with practical deployment considerations. 

While the vulnerability of cycle rickshaw pullers to extreme heat is well recognized, little effort has been devoted to modeling how their physiological biomarkers respond under such conditions. In this study, we collect real-time weather and physiological data using a wearable computing platform from 100 rickshaw pullers in Dhaka, Bangladesh. In parallel, we interview 12 additional rickshaw pullers to explore their knowledge, perceptions, and experiences related to climate change. We propose a Linear Gaussian Bayesian Network (LGBN)-based regression model that predicts key physiological biomarkers based on activity, weather, and demographic features. The model achieves normalized mean absolute error (NMAE) of 0.82, 0.47, 0.65, and 0.67, respectively, for the biomarker: skin temperature, relative cardiac cost, skin conductance response, and skin conductance level. Using climate model projections from 18 CMIP6 global climate models, we layer the LGBN on top of future climate forecasts to conduct a survivability analysis for both current (2023-2025) and future years (2026-2100). Based on the criteriaT_WEGT> 31.1° C andT_skiin>35°C, the analysis shows that a significant percentage of rickshaw pullers (32%) are already facing a high risk of heat-related illness or prolonged exposure to extreme heat(T_WBGT>31.1°C) during regular work hours. In future years, e.g., 2026–2030, based on the CMIP6-based climate models, this percentage can rise to 37 ±17% with an exposure duration of 11.9 ±2 minutes (68% of the trip duration) on average. A similar trend is found based on rickshaw pullers' skin temperature with exposure(T_skin>35°C) durations expanding from 11 minutes (64% of the trip duration) to 13 ± 2 minutes (73% of the trip duration) by 2026-2030. Finally, a Thematic Analysis of interview data provides qualitative insights that complement the current observation and model's predictions in the future. The findings reveal that rickshaw 

Inverse Synthetic Aperture Radar (ISAR) imaging presents a formidable challenge when it comes to small everyday objects due to their limited Radar Cross-Section (RCS) and the inherent resolution constraints of radar systems. Existing ISAR reconstruction methods including backprojection (BP) often require complex setups and controlled environments, rendering them impractical for many real-world noisy scenarios. In this paper, we propose a novel Analysis-through-Synthesis (ATS) framework enabled by Neural Radiance Fields (NeRF) for high-resolution coherent ISAR imaging of small objects using sparse and noisy Ultra-Wideband (UWB) radar data with an inexpensive and portable setup. Our end-to-end framework integrates ultra-wideband radar wave propagation, reflection characteristics, and scene priors, enabling efficient 2D scene reconstruction without the need for costly anechoic chambers or complex measurement test beds. With qualitative and quantitative comparisons, we demonstrate that the proposed method outperforms traditional techniques and generates ISAR images of complex scenes with multiple targets and complex structures in Non-Line-of-Sight (NLOS) and noisy scenarios, particularly with limited number of views and sparse UWB radar scans. This work represents a significant step towards practical, costeffective ISAR imaging of small everyday objects, with broad implications for robotics and mobile sensing applications. 

There has been growing interest in developing ubiquitous technologies to analyze adult-child speech in naturalistic settings such as free play in order to support children's social and academic development, language acquisition, and parent-child interactions. However, these technologies often rely on off-the-shelf speech processing tools that have not been evaluated on child speech or child-directed adult speech, whose unique characteristics might result in significant performance gaps when using models trained on adult speech. This work introduces the Playlogue dataset containing over 33 hours of long-form, naturalistic, play-based adult-child conversations from three different corpora of preschool-aged children. Playlogue enables researchers to train and evaluate speaker diarization and automatic speech recognition models on child-centered speech. We demonstrate the lack of generalizability of existing state-of-the-art models when evaluated on Playlogue, and show how fine-tuning models on adult-child speech mitigates the performance gap to some extent but still leaves considerable room for improvement. We further annotate over 5 hours of the Playlogue dataset with 8668 validated adult and child speech act labels, which can be used to train and evaluate models to provide clinically relevant feedback on parent-child interactions. We investigate the performance of state-of-the-art language models at automatically predicting these speech act labels, achieving significant accuracy with simple chain-of-thought prompting or minimal fine-tuning. We use inhome pilot data to validate the generalizability of models trained on Playlogue, demonstrating its utility in improving speech and language technologies for child-centered conversations. The Playlogue dataset is available for download at https://huggingface.co/datasets/playlogue/playlogue-v1. 

Long-term and high-dose prescription opioid use places individuals at risk for opioid misuse, opioid use disorder (OUD), and overdose. Existing methods for monitoring opioid use and detecting misuse rely on self-reports, which are prone to reporting bias, and toxicology testing, which may be infeasible in outpatient settings. Although wearable technologies for monitoring day-to-day health metrics have gained significant traction in recent years due to their ease of use, flexibility, and advancements in sensor technology, their application within the opioid use space remains underexplored. In the current work, we demonstrate that oral opioid administrations can be detected using physiological signals collected from a wrist sensor. More importantly, we show that models informed by opioid pharmacokinetics increase reliability in predicting the timing of opioid administrations. Forty-two individuals who were prescribed opioids as a part of their medical treatment in-hospital and after discharge were enrolled. Participants wore a wrist sensor throughout the study, while opioid administrations were tracked using electronic medical records and self-reports. We collected 1,983 hours of sensor data containing 187 opioid administrations from the inpatient setting and 927 hours of sensor data containing 40 opioid administrations from the outpatient setting. We demonstrate that a self-supervised pre-trained model, capable of learning the canonical time series of plasma concentration of the drug derived from opioid pharmacokinetics, can reliably detect opioid administration in both settings. Our work suggests the potential of pharmacokinetic-informed, data-driven models to objectively detect opioid use in daily life.