Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “deep potential”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 145 records · Page 8
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Development of interatomic potential for Al–Tb alloys using a deep neural network learning method

An interatomic potential for the Al–Tb alloy around the composition of Al 90 Tb 10 is developed using the deep neural network (DNN) learning method. The atomic configurations and the corresponding total potential energies and forces on each atom obtained from ab initio molecular dynamics (AIMD) simulations are collected to train a DNN model to construct the interatomic potential for the Al–Tb alloy. Here we show that the obtained DNN model can well reproduce the energies and forces calculated by AIMD simulations. Molecular dynamics (MD) simulations using the DNN interatomic potential also accurately describe the structural properties of the Al 90 Tb 10 liquid, such as partial pair correlation functions (PPCFs) and bond angle distributions, in comparison with the results from AIMD simulations. Furthermore, the developed DNN interatomic potential predicts the formation energies of the crystalline phases of the Al–Tb system with an accuracy comparable to ab initio calculations. The structure factors of the Al 90 Tb 10 metallic liquid and glass obtained by MD simulations using the developed DNN interatomic potential are also in good agreement with the experimental X-ray diffraction data. The development of short-range order (SRO) in the Al 90 Tb 10 liquid and the undercooled liquid is also analyzed and three dominant SROs, i.e., Al-centered distorted icosahedron (DISICO) and Tb-centered ‘3661’ and ‘15551’ clusters, respectively, are identified.

36 MATERIALS SCIENCE↗
DOE OSTI2018

Less is more: Sampling chemical space with active learning

We present the development of accurate and transferable machine learning (ML) potentials for predicting molecular energetics is a challenging task. The process of data generation to train such ML potentials is a task neither well understood nor researched in detail. In this work, we present a fully automated approach for the generation of datasets with the intent of training universal ML potentials. It is based on the concept of active learning (AL) via Query by Committee (QBC), which uses the disagreement between an ensemble of ML potentials to infer the reliability of the ensemble’s prediction. QBC allows the presented AL algorithm to automatically sample regions of chemical space where the ML potential fails to accurately predict the potential energy. AL improves the overall fitness of ANAKIN-ME (ANI) deep learning potentials in rigorous test cases by mitigating human biases in deciding what new training data to use. AL also reduces the training set size to a fraction of the data required when using naive random sampling techniques. To provide validation of our AL approach, we develop the COmprehensive Machine-learning Potential (COMP6) benchmark (publicly available on GitHub) which contains a diverse set of organic molecules. Active learning-based ANI potentials outperform the original random sampled ANI-1 potential with only 10% of the data, while the final active learning-based model vastly outperforms ANI-1 on the COMP6 benchmark after training to only 25% of the data. Finally, we show that our proposed AL technique develops a universal ANI potential (ANI-1x) that provides accurate energy and force predictions on the entire COMP6 benchmark. Finally, this universal ML potential achieves a level of accuracy on par with the best ML potentials for single molecules or materials, while remaining applicable to the general class of organic molecules composed of the elements CHNO.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2019

Application of a Deep Learning Network to 3D Modeling of Nuclear Facilities - 19614

The ultimate goal of this study is to minimize (or eliminate) human intervention in the scanning and 3D CAD modeling process in the generation of 3D as-built model of nuclear facilities. This paper describes the application of the state-of-the-art deep learning technology in 3D point cloud data segmentation which is a part of the 3D modeling process. The baseline deep learning network chosen in this study is PointNet. We compare the deep learning approach with a conventional filtering approach, and show the feasibility and potential of deep learning technology. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗
DOE OSTI2016

UV DRIVEN EVAPORATION OF CLOSE-IN PLANETS: ENERGY-LIMITED, RECOMBINATION-LIMITED, AND PHOTON-LIMITED FLOWS

We have investigated the evaporation of close-in exoplanets irradiated by ionizing photons. We find that the properties of the flow are controlled by the ratio of the recombination time to the flow timescale. When the recombination timescale is short compared to the flow timescale, the flow is in approximate local ionization equilibrium with a thin ionization front where the photon mean free path is short compared to the flow scale. In this “recombination-limited” flow the mass-loss scales roughly with the square root of the incident flux. When the recombination time is long compared to the flow timescale the ionization front becomes thick and encompasses the entire flow with the mass-loss rate scaling linearly with flux. If the planet's potential is deep, then the flow is approximately “energy-limited”; however, if the planet's potential is shallow, then we identify a new limiting mass-loss regime, which we term “photon-limited.” In this scenario, the mass-loss rate is purely limited by the incoming flux of ionizing photons. We have developed a new numerical approach that takes into account the frequency dependence of the incoming ionizing spectrum and performed a large suite of 1D simulations to characterize UV driven mass-loss around low-mass planets. We find that the flow is “recombination-limited” at high fluxes but becomes “energy-limited” at low fluxes; however, the transition is broad occurring over several orders of magnitude in flux. Finally, we point out that the transitions between the different flow types do not occur at a single flux value but depend on the planet's properties, with higher-mass planets becoming “energy-limited” at lower fluxes.

79 ASTRONOMY AND ASTROPHYSICS↗
NASA NTRSDate not supplied

Imaging Science Requirements for a Uranus Flagship Mission

Introduction: Our presentation will summarize the requirements for UV/Visible/Near-IR remote-sensing imaging science instruments (including mapping spectrometers) envisioned for the Uranus Orbiter and Probe (UOP) mission recommended by the recent planetary decadal survey. The Uranus system offers many targets for imaging investigations. Imaging science investigations are crucial to answering fundamental questions about various physical processes that shape the Uranian system and drive its evolution. Each science objective addressed through imaging has requirements for observation and illumination geometries as well as coverage and resolution in the spatial, spectral and temporal dimensions. By combining the requirements for the diverse science targets in the Uranian system, we present a set of overarching imaging science requirements for UOP. Atmospheric Science: The study of Uranus’ atmosphere is crucial in understanding the planet's energy balance and interior heat flux. Imaging observations map how the atmosphere scatters incident solar radiation. Imaging is key to characterizing planetary-scale energy/material transport though global atmospheric dynamics and local meteorology. Hyperspectral mapping combined with radiative transfer analysis reveals the vertical distribution of clouds and hazes. Chronicling temporal variabilities of atmospheric features is key to understanding the various processes that operate within the dynamic Uranian atmosphere. A comprehensive understanding of Uranus' atmosphere is essential in deciphering the planet's origins and evolution as well as providing insight into the broader field of planetary and exoplanetary science. Magnetospheric Interactions: Imaging science is critical to understand the elusive Uranian aurorae. Mapping the aurorae and their temporal dynamics will reveal the interaction of Uranus’ complex magnetosphere with the solar wind. The apparent decades-long cooling of Uranus’ upper atmosphere is thought to be due to changes in energy input from the overlying magnetosphere, and so mapping and quantifying the auroral evidence for this energy flux is particularly relevant for upper atmospheric science In addition, Uranus’ large moons may generate auroral footprints, which can be used to constrain the planet’s unusual magnetic field structure. Satellite Interiors: Imaging science can address the potential habitability of the larger moons of Uranus by measuring their shapes and rotational states (including librations), which combined with gravity data are crucial to understanding their internal density structure. Satellite Origins: Origin markers such as organics and volatiles may be preserved on the satellite surfaces, which may be sensed by UV/IR mapping. Imaging combined with the moons’ densities may reveal whether the moons formed in Uranus’ circumplanetary disk, rings, or from ejecta of a giant impact on Uranus. Satellite Geology: Imaging investigations, including stereo imaging to obtain topography, will enable constraining each satellite’s geologic history by investigating the distributions and morphologies of impact craters, tectonic and potential cryovolcanic features, and regolith and mass wasting deposits. These records will help unravel the past complex history of orbital resonances that occurred between Uranian satellites, which may have driven geologic activity. Analyses of these features will also point to resurfacing styles and modification processes important for understanding each moon’s thermal evolution, and any present-day geological activity, including potential subsurface deep oceans or possible current activity. Additionally, hyperspectral imaging will enable mapping the surface composition and identifying the potential presence of material that may point to a subsurface ocean, such as ammonia-bearing species. Rings and Small Moons: Imaging observations elucidate the physical processes that contribute to the complex ring-moon system’s inferred unstable state. Imaging surveys will seek smaller moons and other debris that could provide evidence of past collisions, while precise astrometric data of moons will determine the current dynamical state of the system. Many of Uranus' rings are exceptionally narrow, and images of the rings covering a broad range of longitudes, and high-resolution images of features within the rings will reveal how these rings are confined and maintained, as well as illustrating the rings' internal structure and dynamics. Images of the rings and moons obtained at a variety of wavelengths and lighting conditions are also needed to quantify the composition and the size distribution of the small satellites and ring particles.

K M Sayanagi↗
DOE OSTI2017

Deep Learning for Automated Extraction of Primary Sites from Cancer Pathology Reports

Pathology reports are a primary source of information for cancer registries which process high volumes of free-text reports annually. Information extraction and coding is a manual, labor-intensive process. Here in this study we investigated deep learning and a convolutional neural network (CNN), for extracting ICDO- 3 topographic codes from a corpus of breast and lung cancer pathology reports. We performed two experiments, using a CNN and a more conventional term frequency vector approach, to assess the effects of class prevalence and inter-class transfer learning. The experiments were based on a set of 942 pathology reports with human expert annotations as the gold standard. CNN performance was compared against a more conventional term frequency vector space approach. We observed that the deep learning models consistently outperformed the conventional approaches in the class prevalence experiment, resulting in micro and macro-F score increases of up to 0.132 and 0.226 respectively when class labels were well populated. Specifically, the best performing CNN achieved a micro-F score of 0.722 over 12 ICD-O-3 topography codes. Transfer learning provided a consistent but modest performance boost for the deep learning methods but trends were contingent on CNN method and cancer site. Finally, these encouraging results demonstrate the potential of deep learning for automated abstraction of pathology reports.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2017

Background rejection in NEXT using deep neural networks

Here, we investigate the potential of using deep learning techniques to reject background events in searches for neutrinoless double beta decay with high pressure xenon time projection chambers capable of detailed track reconstruction. The differences in the topological signatures of background and signal events can be learned by deep neural networks via training over many thousands of events. These networks can then be used to classify further events as signal or background, providing an additional background rejection factor at an acceptable loss of efficiency. The networks trained in this study performed better than previous methods developed based on the use of the same topological signatures by a factor of 1.2 to 1.6, and there is potential for further improvement.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
NASA NTRS2012

Evaluation of Deep-Space Laser Communication under Different Mission Scenarios

A number of space agencies, including NASA, are considering free-space laser communications as a means for returning higher data-rates from future space missions. In this paper, potential deep-space missions are evaluated to show that with optical communication a 10x increase relative to state-of-the art telecommunication systems could be achieved. The maximum deep-space distance where ground transmitted laser beacons could assist acquisition and tracking; and operating points where optical communication performance degrades faster than the inverse square distance are also discussed.

mission scenarios↗
DOE OSTI2018

The impact of baryonic discs on the shapes and profiles of self-interacting dark matter haloes

We employ isolated N-body simulations to study the response of self-interacting dark matter (SIDM) haloes in the presence of the baryonic potentials. Dark matter self-interactions lead to kinematic thermalization in the inner halo, resulting in a tight correlation between the dark matter and baryon distributions. A deep baryonic potential shortens the phase of SIDM core expansion and triggers core contraction. This effect can be further enhanced by a large self-scattering cross-section. We find the final SIDM density profile is sensitive to the baryonic concentration and the strength of dark matter self-interactions. Assuming a spherical initial halo, we also study evolution of the SIDM halo shape together with the density profile. The halo shape at later epochs deviates from spherical symmetry due to the influence of the non-spherical disc potential, and its significance depends on the baryonic contribution to the total gravitational potential, relative to the dark matter one. In addition, we construct a multicomponent model for the Milky Way, including an SIDM halo, a stellar disc, and a bulge, and show it is consistent with observations from stellar kinematics and streams.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Characterizing Structure-Dependent TiS 2 /Water Interfaces Using Deep-Neural-Network-Assisted Molecular Dynamics

As a promising layered electrode material, TiS 2 -based capacitive deionization (CDI) devices for water desalination have attracted significant attention. However, TiS 2 /H 2 O interfacial features, potentially important for device optimization, remain unidentified. Using Deep Potential Molecular Dynamics (DPMD), we characterized distinct aqueous interfaces introduced by four TiS 2 terminations expected to be present as water intercalates into TiS 2 , namely, Armchair, Zigzag, Zigzag-L, and Zigzag-R. First, we assessed important representative physical properties of the system to validate the deep potentials (DPs). DPMD simulations agree well with experiments and first-principles simulations, suggesting the DPs are accurate and reliable. Subsequent simulations of these TiS 2 /water interfaces revealed how TiS 2 surface termination influences the structure of interfacial water. This effect is most evident in the first and second water layers close to the TiS 2 surface, and more pronounced when spontaneous dissociative adsorption of water occurs. The extent of water dissociation on each surface was evaluated using enhanced sampling. Zigzag-L is the only interface where proton transfer from adsorbed water to TiS 2 surface S atoms is thermodynamically and kinetically favored. The coexistence of surface four-fold-coordinated Ti (Ti 4c ) and one-fold-coordinated S (S 1c ) is found to be essential to making proton transfer feasible on the Zigzag-L surface. Furthermore, remaining unprotonated S 1c atoms can act as good proton acceptors after water dissociation. Thus, TiS 2 with Zigzag-L termination may be a surface to avoid in CDI device construction, given that pH fluctuations adversely affect performance. Furthermore, this work provides new understanding of TiS 2 /H 2 O interfacial features that could aid future design and optimization of TiS 2 -based CDI devices for water desalination.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Triangle Method for Dense ReLU Layers [SWR-25-72]

This software is an implementation of the methods for initializing and training neural networks to be more efficient per parameter, described more fully below and in the related publication: In theory, depth should make a ReLU network EXPONENTIALLY more efficient by enabling it to produce an exponential number of piecewise linear sections in its output. This reasoning is largely based on the work of mathematicians that have hand-constructed networks that make good use of depth. In practice however, even very deep ReLU networks that have been randomly initialized will behave identically to their shallow counterparts - missing an entire exponential dimension of efficiency. The triangle method is a first attempt at realizing the exponential potential of deep networks. Instead of randomly setting weights, we force pairs of neurons in each layer learn to build triangles (i.e. functions from [0,1] -> [0,1] that look like triangles). This is a very efficient pattern for generating lots of linear pieces because composing two triangular functions doubles the number of pieces with each composition. The triangle method is more than just a different initialization, it is a new paradigm of training. Instead of making direct updates to the matrix weights, we do an extra step of backpropagation to collect the derivatives of the loss function with respect to the shapes of the triangles, training them to tilt left or right. This process essentially holds the networks hand throughout the loss landscape and forces it to always use depth effectively by producing triangular shapes internally. This can produce several orders of magnitude of improvement on convex one-dimensional regression problems. Much more theoretical work is needed to realize its full potential beyond this context, but the implementation in this repository will still work in arbitrary numbers of dimensions. The file Triangle_Method.py is a generalized form of the method that will build each neuron its own custom 1-d convex activation function (with exponential efficiency). Example usage on one dimensional problems can be found in Example_Usage.ipynb and an example of using this in a real neural network can be found in Example_VGG16_CIFAR10.ipynb.

Milkert, Max [National Renewable Energy Laboratory↗
DOE OSTI2019

DeepCMB: Lensing reconstruction of the cosmic microwave background with deep neural networks

Next-generation cosmic microwave background (CMB) experiments will have lower noise and therefore increased sensitivity, enabling improved constraints on fundamental physics parameters such as the sum of neutrino masses and the tensor-to-scalar ratio r. Achieving competitive constraints on these parameters requires high signal-to-noise extraction of the projected gravitational potential from the CMB maps. Standard methods for reconstructing the lensing potential employ the quadratic estimator (QE). However, the QE performs suboptimally at the low noise levels expected in upcoming experiments. Other methods, like maximum likelihood estimators (MLE), are under active development. In this work, we demonstrate reconstruction of the CMB lensing potential with deep convolutional neural networks (CNN) - ie, a ResUNet. The network is trained and tested on simulated data, and otherwise has no physical parametrization related to the physical processes of the CMB and gravitational lensing. We show that, over a wide range of angular scales, ResUNets recover the input gravitational potential with a higher signal-to-noise ratio than the QE method, reaching levels comparable to analytic approximations of MLE methods. We demonstrate that the network outputs quantifiably different lensing maps when given input CMB maps generated with different cosmologies. We also show we can use the reconstructed lensing map for cosmological parameter estimation. This application of CNN provides a few innovations at the intersection of cosmology and machine learning. First, while training and regressing on images, we predict a continuous-variable field rather than discrete classes. Second, we are able to establish uncertainty measures for the network output that are analogous to standard methods. We expect this approach to excel in capturing hard-to-model non-Gaussian astrophysical foreground and noise contributions.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2019

DeepCMB: Lensing Reconstruction of the CMB with Deep Neural Networks

Next-generation cosmic microwave background (CMB) experiments will have lower noise and therefore increased sensitivity, enabling improved constraints on fundamental physics parameters such as the sum of neutrino masses and the tensor-to-scalar ratio r. Achieving competitive constraints on these parameters requires high signal-to-noise extraction of the projected gravitational potential from the CMB maps. Standard methods for reconstructing the lensing potential employ the quadratic estimator (QE). However, the QE performs suboptimally at the low noise levels expected in upcoming experiments. Other methods, like maximum likelihood estimators (MLE), are under active development. In this work, we demonstrate reconstruction of the CMB lensing potential with deep convolutional neural networks (CNN) - ie, a ResUNet. The network is trained and tested on simulated data, and otherwise has no physical parametrization related to the physical processes of the CMB and gravitati onal lensing. We show that, over a wide range of angular scales, ResUNets recover the input gravitational potential with a higher signal-to-noise ratio than the QE method, reaching levels comparable to analytic approximations of MLE methods. We demonstrate that the network outputs quantifiably different lensing maps when given input CMB maps generated with different cosmologies. We also show we can use the reconstructed lensing map for cosmological parameter estimation. This application of CNN provides a few innovations at the intersection of cosmology and machine learning. First, while training and regressing on images, we predict a continuous-variable field rather than discrete classes. Second, we are able to establish uncertainty measures for the network output that are analogous to standard methods. We expect this approach to excel in capturing hard-to-model non-Gaussian astrophysical foreground and noise contributions.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2016

TH-CD-206-05: Machine-Learning Based Segmentation of Organs at Risks for Head and Neck Radiotherapy Planning

Purpose: Accurate and efficient delineation of tumor target and organs-at-risks is essential for the success of radiotherapy. In reality, despite of decades of intense research efforts, auto-segmentation has not yet become clinical practice. In this study, we present, for the first time, a deep learning-based classification algorithm for autonomous segmentation in head and neck (HaN) treatment planning. Methods: Fifteen HN datasets of CT, MR and PET images with manual annotation of organs-at-risk (OARs) including spinal cord, brainstem, optic nerves, chiasm, eyes, mandible, tongue, parotid glands were collected and saved in a library of plans. We also have ten super-resolution MR images of the tongue area, where the genioglossus and inferior longitudinalis tongue muscles are defined as organs of interest. We applied the concepts of random forest- and deep learning-based object classification for automated image annotation with the aim of using machine learning to facilitate head and neck radiotherapy planning process. In this new paradigm of segmentation, random forests were used for landmark-assisted segmentation of super-resolution MR images. Alternatively to auto-segmentation with random forest-based landmark detection, deep convolutional neural networks were developed for voxel-wise segmentation of OARs in single and multi-modal images. The network consisted of three pairs of convolution and pooing layer, one RuLU layer and a softmax layer. Results: We present a comprehensive study on using machine learning concepts for auto-segmentation of OARs and tongue muscles for the HaN radiotherapy planning. An accuracy of 81.8% in terms of Dice coefficient was achieved for segmentation of genioglossus and inferior longitudinalis tongue muscles. Preliminary results of OARs regimentation also indicate that deep-learning afforded an unprecedented opportunities to improve the accuracy and robustness of radiotherapy planning. Conclusion: A novel machine learning framework has been developed for image annotation and structure segmentation. Our results indicate the great potential of deep learning in radiotherapy treatment planning.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2021

Quantum reactive scattering calculations for the cold and ultracold Li + LiNa → Li 2 + Na reaction

A first-principles based quantum dynamics study of the Li + LiNa(ν = 0, j = 0) → Li 2 (ν', j') + Na reaction is reported for collision energies spanning the ultracold (1 nK) to cold (1 K) regimes. A full-dimensional ab initio potential energy surface for the ground electronic state of Li 2 Na is utilized that includes an accurate treatment of the long-range interactions. The Li + LiNa reaction is barrierless and exoergic and exhibits a deep attractive potential well that supports complex formation. Thus, significant reactivity occurs even for collision temperatures approaching absolute zero. The reactive scattering calculations are based on a numerically exact time-independent quantum dynamics methodology in hyperspherical coordinates. Total and rotationally resolved rate coefficients are reported at 56 collision energies and include all contributing partial waves. Several shape resonances are observed in many of the rotationally resolved rate coefficients and a small resonance feature is also reported in the total rate coefficient near 50 mK. Furthermore, of particular interest, the angular distributions or differential cross sections are reported as a function of both the collision energy and scattering angle. Unique quantum fingerprints (bumps, channels, and ripples) are observed in the angular distributions for each product rotational state due to quantum interference and shape resonance contributions. The Li + LiNa reaction is under active experimental investigation so that these intriguing features could be verified experimentally when sufficient product state resolution becomes feasible for collision energies below 1 K.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2023

Developing machine-learning potentials to study properties of the tritium formation and diffusivity in pure and defective Zircaloy-4 getters

Objective of this work was to study the formation of Sn impurity and hydride phases in Zr and study their impacts on the diffusion kinetics of 3 H by using DFT based ML approach. By implementing deep neural potential (DNP) technique, we developed potential for Zr-H and Sn impurity systems and validated the DNP by using DFT results. We found that by implementing machine learning approach, it is possible to achieve accuracy comparable to DFT level for the more realistic models by using less computational time and resources.

36 MATERIALS SCIENCE↗
DOE OSTI2020

Achieving large uniform tensile elasticity in microfabricated diamond

Diamond is not only the hardest material in nature, but is also an extreme electronic material with an ultrawide bandgap, exceptional carrier mobilities, and thermal conductivity. Straining diamond can push such extreme figures of merit for device applications. We microfabricated single-crystalline diamond bridge structures with ~1 micrometer length by ~100 nanometer width and achieved sample-wide uniform elastic strains under uniaxial tensile loading along the [100], [101], and [111] directions at room temperature. We also demonstrated deep elastic straining of diamond microbridge arrays. The ultralarge, highly controllable elastic strains can fundamentally change the bulk band structures of diamond, including a substantial calculated bandgap reduction as much as ~2 electron volts. Our demonstration highlights the immense application potential of deep elastic strain engineering for photonics, electronics, and quantum information technologies.

Dang, Chaoqun↗
DOE OSTI2025

Applying Deep Learning for Wildfire Identification: Economical and Accessible Solutions Leveraging Small Datasets

Wildfires significantly impact human health, air quality, visibility, weather, and climate change and cause substantial economic losses. While state and county-operated air quality monitors provide critical insights during wildfires, they are not available in all regions. This highlights the need for affordable, accessible tools that allow the general public to assess air quality impacts. In this study, we apply machine learning with deep neural networks to diagnose air quality rapidly from sky images taken at the Pacific Northwest National Laboratory in Richland, WA, USA. Using a convolutional neural network (CNN) framework, we trained a deep learning model to classify air quality indices based on sky images. By leveraging transfer learning, our approach fine-tunes a pre-trained model on a small dataset of sky images, significantly reducing training time while maintaining high accuracy. Our results demonstrate the potential of deep learning to provide rapid air quality diagnostics during wildfire episodes, offering early warnings to the public and enabling timely mitigation strategies, particularly for vulnerable populations. Additionally, we show that lower respiratory infections pose the highest health risk during acute smoke exposures. Reactive oxygen species (ROS) from wildfire particles further exacerbate health risks by triggering inflammation and other adverse effects.

54 ENVIRONMENTAL SCIENCES↗