Recognizing food types through sensor signals for unseen users remains remarkably challenging, despite extensive recent studies. The efficacy of prior machine learning techniques is dwarfed by giant variations of data collected from multiple participants, partly because users have varied chewing habits and wear sensor devices in various manners. This work treats the problem as an instance of the domain adaptation problem, where each user represents a domain. We develop the first multi-source domain adaptation (MSDA) method for food-typing recognition, which consists of three major components: stratified normalization, a multi-source domain adaptor, and adaptive ensemble learning. New techniques are developed for each component. Using a real-world dataset comprised of 15 participants, we demonstrate that our method achieves
Note: When clicking on a Digital Object Identifier (DOI) number, you will be taken to an external site maintained by the publisher.
Some full text articles may not yet be available without a charge during the embargo (administrative interval).
What is a DOI Number?
Some links on this page may take you to non-federal websites. Their policies may differ from this site.
-
to\(1.33\times\) improvement in accuracy compared with nine state-of-the-art MSDA baselines. Additionally, we perform an in-depth ablation study to examine the behavior of each component and confirm their efficacy.\(2.13\times\) Free, publicly-accessible full text available September 23, 2025 -
Earth’s inner core is predominantly composed of solid iron (Fe) and displays intriguing properties such as strong shear softening and an ultrahigh Poisson’s ratio. Insofar, physical mechanisms to explain these features coherently remain highly debated. Here, we have studied longitudinal and shear wave velocities of hcp-Fe (hexagonal close-packed iron) at relevant pressure–temperature conditions of the inner core using in situ shock experiments and machine learning molecular dynamics (MLMD) simulations. Our results demonstrate that the shear wave velocity of hcp-Fe along the Hugoniot in the premelting condition, defined as
T/Tm (Tm : melting temperature of iron) above 0.96, is significantly reduced by ~30%, while Poisson’s ratio jumps to approximately 0.44. MLMD simulations at 230 to 330 GPa indicate that collective motion with fast diffusive atomic migration occurs in premelting hcp-Fe primarily along [100] or [010] crystallographic direction, contributing to its elastic softening and enhanced Poisson’s ratio. Our study reveals that hcp-Fe atoms can diffusively migrate to neighboring positions, forming open-loop and close-loop clusters in the inner core conditions. Hcp-Fe with collective motion at the inner core conditions is thus not an ideal solid previously believed. The premelting hcp-Fe with collective motion behaves like an extremely soft solid with an ultralow shear modulus and an ultrahigh Poisson’s ratio that are consistent with seismic observations of the region. Our findings indicate that premelting hcp-Fe with fast diffusive motion represents the underlying physical mechanism to help explain the unique seismic and geodynamic features of the inner core. -
Automatic food type recognition is an essential task of dietary monitoring. It helps medical professionals recognize a user’s food contents, estimate the amount of energy intake, and design a personalized intervention model to prevent many chronic diseases, such as obesity and heart disease. Various wearable and mobile devices are utilized as platforms for food type recognition. However, none of them has been widely used in our daily lives and, at the same time, socially acceptable enough for continuous wear. In this paper, we propose a food type recognition method that takes advantage of Airpods Pro, a pair of widely used wireless in-ear headphones designed by Apple, to recognize 20 different types of food. As far as we know, we are the first to use this socially acceptable commercial product to recognize food types. Audio and motion sensor data are collected from Airpods Pro. Then 135 representative features are extracted and selected to construct the recognition model using the lightGBM algorithm. A real-world data collection is conducted to comprehensively evaluate the performance of the proposed method for seven human subjects. The results show that the average f1-score reaches 94.4% for the ten-fold cross- validation test and 96.0% for the self-evaluation test.more » « less
-
Abstract. Nitrous oxide (N2O) is a long-lived potent greenhouse gas and stratospheric ozone-depleting substance that has been accumulating in the atmosphere since the preindustrial period. The mole fraction of atmospheric N2O has increased by nearly 25 % from 270 ppb (parts per billion) in 1750 to 336 ppb in 2022, with the fastest annual growth rate since 1980 of more than 1.3 ppb yr−1 in both 2020 and 2021. According to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC AR6), the relative contribution of N2O to the total enhanced effective radiative forcing of greenhouse gases was 6.4 % for 1750–2022. As a core component of our global greenhouse gas assessments coordinated by the Global Carbon Project (GCP), our global N2O budget incorporates both natural and anthropogenic sources and sinks and accounts for the interactions between nitrogen additions and the biogeochemical processes that control N2O emissions. We use bottom-up (BU: inventory, statistical extrapolation of flux measurements, and process-based land and ocean modeling) and top-down (TD: atmospheric measurement-based inversion) approaches. We provide a comprehensive quantification of global N2O sources and sinks in 21 natural and anthropogenic categories in 18 regions between 1980 and 2020. We estimate that total annual anthropogenic N2O emissions have increased 40 % (or 1.9 Tg N yr−1) in the past 4 decades (1980–2020). Direct agricultural emissions in 2020 (3.9 Tg N yr−1, best estimate) represent the large majority of anthropogenic emissions, followed by other direct anthropogenic sources, including fossil fuel and industry, waste and wastewater, and biomass burning (2.1 Tg N yr−1), and indirect anthropogenic sources (1.3 Tg N yr−1) . For the year 2020, our best estimate of total BU emissions for natural and anthropogenic sources was 18.5 (lower–upper bounds: 10.6–27.0) Tg N yr−1, close to our TD estimate of 17.0 (16.6–17.4) Tg N yr−1. For the 2010–2019 period, the annual BU decadal-average emissions for both natural and anthropogenic sources were 18.2 (10.6–25.9) Tg N yr−1 and TD emissions were 17.4 (15.8–19.20) Tg N yr−1. The once top emitter Europe has reduced its emissions by 31 % since the 1980s, while those of emerging economies have grown, making China the top emitter since the 2010s. The observed atmospheric N2O concentrations in recent years have exceeded projected levels under all scenarios in the Coupled Model Intercomparison Project Phase 6 (CMIP6), underscoring the importance of reducing anthropogenic N2O emissions. To evaluate mitigation efforts and contribute to the Global Stocktake of the United Nations Framework Convention on Climate Change, we propose the establishment of a global network for monitoring and modeling N2O from the surface through to the stratosphere. The data presented in this work can be downloaded from https://doi.org/10.18160/RQ8P-2Z4R (Tian et al., 2023).
-
Abstract Formation of thick, high energy density, flexible solid supercapacitors is challenging because of difficulties infilling gel electrolytes into porous electrodes. Incomplete infilling results in a low capacitance and poor mechanical properties. Here we report a bottom-up infilling method to overcome these challenges. Electrodes up to 500 μm thick, formed from multi-walled carbon nanotubes and a composite of poly(3,4-ethylenedioxythiophene), polystyrene sulfonate and multi-walled carbon nanotubes are successfully infilled with a polyvinyl alcohol/phosphoric acid gel electrolyte. The exceptional mechanical properties of the multi-walled carbon nanotube-based electrode enable it to be rolled into a radius of curvature as small as 0.5 mm without cracking and retain 95% of its initial capacitance after 5000 bending cycles. The areal capacitance of our 500 μm thick poly(3,4-ethylenedioxythiophene), polystyrene sulfonate, multi-walled carbon nanotube-based flexible solid supercapacitor is 2662 mF cm–2at 2 mV s–1, at least five times greater than current flexible supercapacitors.