Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Distance learning”

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

Off-Angle Iris Correction Methods

In many real world iris recognition systems obtaining consistent frontal images is problematic do to inexperienced or uncooperative users, untrained operators, or distracting environments. As a result many collected images are unusable by modern iris matchers. In this chapter we present four methods for correcting off-angle iris images to appear frontal which makes them compatible with existing iris matchers. The methods include an affine correction, a retraced model of the human eye, measured displacements, and a genetic algorithm optimized correction. The affine correction represents a simple way to create an iris image that appears frontal but it does not account for refractive distortions of the cornea. The other method account for refraction. The retraced model simulates the optical properties of the cornea. The other two methods are data driven. The first uses optical flow to measure the displacements of the iris texture when compared to frontal images of the same subject. The second uses a genetic algorithm to learn a mapping that optimizes the Hamming Distance scores between off-angle and frontal images. In this paper we hypothesize that the biological model presented in our earlier work does not adequately account for all variations in eye anatomy and therefore the two data-driven approaches should yield better performance. Results are presented using the commercial VeriEye matcher that show that the genetic algorithm method clearly improves over prior work and makes iris recognition possible up to 50 degrees off-angle.

iris recognition↗

Technical Performance and Cost Optimization of Unobtrusive Multi-static Serial LiDAR Imager (UMSLI) for Wide-area Surveillance and Identification of Marine Life at Marine Energy Installations

Florida Atlantic University developed an underwater optical monitoring system prototype - Unobtrusive Multi-static Serial LiDAR Imager (UMSLI) - suitable for marine energy full project lifecycle observation (baseline, commissioning, and decommissioning), with an automated real-time classification of marine animals. With precursor 2014 DOE funding (award DE-EE0006787), a prototype UMSLI was demonstrated in a controlled laboratory environment and achieved a TRL 6. This EERE DE-EE0007828 award aimed to both increase the TRL of the UMSLI by improving the technology performance (e.g., increase distance of marine animal target detection capability, add additional species classification capabilities, improve the system performance during the day, etc.) and reduce the cost. The UMSLI presents a novel application of underwater distributed Light Detection And Ranging (LiDAR), an advanced remote sensing method that uses light in the form of laser pulses for an application, this paired with an algorithm provides 360 degrees underwater detection, imaging, and classification of marine life. This solution for underwater monitoring of biota preserves the advantages of traditional optical and acoustic solutions while overcoming many associated disadvantages for marine energy site environmental monitoring, such as difficulties in species detection and classification in low light or night and turbid environments. This new approach is a purposefully designed, reconfigurable adaptation of an existing class 3B laser technology into one UMSLI instrument that can be easily mounted on or around different classes of marine energy equipment, such as devices to capture ocean current energy or wave energy. The system uses relatively low average power, utilizes far-red (> 635nm) laser illumination to be invisible and eye-safe to marine animals, is compact, and cost-effective. The equipment is designed for long-term, maintenance-free operations (i.e., current design is targeting more than 7 days continuous operation), to inherently generate a sparse primary dataset that only includes detected anomalies, and to allow robust real-time automated animal classification and identification with a low data bandwidth requirement. The technology’s overarching goal for application, is a system that can be deployed to collect pre-installation baseline species observations at a proposed marine energy deployment site with minimal post-processing overhead. The envisioned system will also produce high-resolution imagery of marine animals through a wide range of conditions and support automated tracking and notification of the presence of managed animals within established perimeters of marine energy equipment to satisfy deployed marine energy projects’ endangered and threatened species monitoring requirements. Through the current project, we demonstrated the UMSLI prototype in an operational environment and increased the UMSLI’s Technology Readiness Level from 6 to 7. The project resulted in many novel technologies, including: 1) an eye-safe red laser based, low-cost LiDAR system that can detect targets up to 10 meters distance; 2) the GAN-based machine learning underwater LiDAR image enhancement technique (this is the first known application of GAN technique in underwater LiDAR); 3) LiDAR-based real-time automated detection capabilities; and 4) a template matching based automated classification tool. These technologies build a solid foundation for future efforts to develop an extended range electro-optical monitoring system suitable for marine energy deployments. In addition to technology development, a key lesson learned is that addressing regulatory and safety requirements must be front and center in any marine energy monitoring applications.

16 TIDAL AND WAVE POWER↗

Allosteric prediction via convolutional neural networks and protein structural and dynamical features

Allostery is the phenomenon whereby a binding event or covalent modification at one site in a protein modulates function at a distal site, thus changing a protein’s functional state. As such, it is a ubiquitous aspect of protein functional regulation. Computationally predicting allosteric states is important as part of the broader challenge of functional annotation, but it also has practical implications for drug development, as targeting an allosteric site often affords greater specificity compared with targeting an orthosteric site. This study introduces a machine learning approach to predict the allosteric functional state using the small G-protein KRas as the model system, due to its implication in many types of cancer and being well studied as a result with many x-ray crystallographic structures of KRas available with different mutations and ligands bound. Using structural and dynamical features that can be cast as images, namely interatomic distances, contact maps, covariance, and mutual information, supervised learning was performed using convolutional neural networks. Two pretrained convolutional neural network architectures, GoogLeNet and ResNet18, were fine-tuned to classify KRas into active or inactive states based on these features. Across training regimes, atomic contact maps emerged as the most effective structural feature, whereas linearized mutual information outperformed covariance in capturing dynamical correlations relevant to allostery. Models achieved significant validation accuracy, with atomic contact maps yielding up to 90% accuracy. In conclusion, the findings suggest that integrating global structural rearrangements and correlated motion patterns with deep learning can reliably predict protein allosteric states, offering a promising framework for understanding allosteric regulation and developing targeted therapeutics.

Rajeshwar T., Rajitha [Oak Ridge National Laborato↗

Predicting Long-Range Traversability from Short-Range Stereo-Derived Geometry

Based only on its appearance in imagery, this program uses close-range 3D terrain analysis to produce training data sufficient to estimate the traversability of terrain beyond 3D sensing range. This approach is called learning from stereo (LFS). In effect, the software transfers knowledge from middle distances, where 3D geometry provides training cues, into the far field where only appearance is available. This is a viable approach because the same obstacle classes, and sometimes the same obstacles, are typically present in the mid-field and the farfield. Learning thus extends the effective look-ahead distance of the sensors.

Turmon, Michael↗

Autonomous Formation Flight: Project Overview

Objectives: a) Map the vortex effects; b) Formation Auto-Pilot Requirements. Two NASA F/A-18 aircraft in formation: a) NASA 845 Systems Research Aircraft; b) NASA 847 Support Aircraft. Flight Conditions: M = 0.56, 25000 feet (Subsonic condition); b) M = 0.86, 36000 feet (Transonic condition). Nose-To-Tail (N2T) Distances: 20, 55, 110 and 190 feet. Lessons learned: a) Controllable flight in vortex is possible with pilot feedback (displays); b) Position hold at best C(sub D), is attainable; c) Best drag location is close to max rolling moment; e) Drag reductions demonstrated up to 22% (WFE up to 20%); f) Induced drag results compare favorably with simple prediction model; g) "Sweet Spot" (lateral & vertical area > 25%) is larger than predicted; h) Larger wing overlaps result in sign reversals in roll, yaw; i) As predicted, favorable effects degrade gradually with increased nose-to-tail distances after peaking at 3 span lengths aft; and j) Demonstrated - over 100 N mi (>15%) range improvement and 650 lbs (14%) fuel savings on actual simulated F/A-18 cruise mission.

Cole, Jennifer↗

Rocket Launch Detection with Smartphone Audio and Transfer Learning

Rocket launches generate infrasound signatures that have been detected at great distances. Due to the sparsity of the networks that have made these detections, however, most signals are detected tens of minutes to hours after the rocket launch. In this work, a method of near-real-time detection of rocket launches using data from a network of smartphones located 10–70 km from launch sites is presented. A machine learning model is trained and tested on the open-access Aggregated Smartphone Timeseries of Rocket-generated Acoustics (ASTRA), Smartphone High-explosive Audio Recordings Dataset (SHAReD), and ESC-50 datasets, resulting in a final accuracy of 97% and a false positive rate of <1%. The performance and behavior of the model are summarized, and its suitability for persistent monitoring applications is discussed.

acoustics↗

Improving underwater localization accuracy with machine learning

Machine learning classification and regression algorithms were applied to calibrate the localization errors of a time-difference-of-arrival (TDOA)-based acoustic sensor array used for tracking salmon passage through a hydroelectric dam on the Snake River, Washington, USA. The locations of stationary and mobile acoustic tags were first tracked using the approximate maximum likelihood algorithm. Next, ensembles of classification trees successfully identified and filtered data points with large localization errors. This prefiltering step allowed the creation of a machine-learned regression model function, which decreased the median distance error by 50% for the stationary tracks and by 34% for the mobile tracks. It also extended the previous range of sub-meter localization accuracy from 100 m to 250 m horizontal distance from the dam face (the receivers). Median distance errors in the depth direction were especially decreased, falling from 0.49 m to 0.04 m in the stationary tracks and from 0.38 m to 0.07 m in the mobile tracks. These methods would have application to the calibration of error in any TDOA-based sensor network with a steady environment and array configuration.

47 OTHER INSTRUMENTATION↗

Multi-head attention-based U-Nets for predicting protein domain boundaries using 1D sequence features and 2D distance maps

Abstract The information about the domain architecture of proteins is useful for studying protein structure and function. However, accurate prediction of protein domain boundaries (i.e., sequence regions separating two domains) from sequence remains a significant challenge. In this work, we develop a deep learning method based on multi-head U-Nets (called DistDom) to predict protein domain boundaries utilizing 1D sequence features and predicted 2D inter-residue distance map as input. The 1D features contain the evolutionary and physicochemical information of protein sequences, whereas the 2D distance map includes the structural information of proteins that was rarely used in domain boundary prediction before. The 1D and 2D features are processed by the 1D and 2D U-Nets respectively to generate hidden features. The hidden features are then used by the multi-head attention to predict the probability of each residue of a protein being in a domain boundary, leveraging both local and global information in the features. The residue-level domain boundary predictions can be used to classify proteins as single-domain or multi-domain proteins. It classifies the CASP14 single-domain and multi-domain targets at the accuracy of 75.9%, 13.28% more accurate than the state-of-the-art method. Tested on the CASP14 multi-domain protein targets with expert annotated domain boundaries, the average per-target F1 measure score of the domain boundary prediction by DistDom is 0.263, 29.56% higher than the state-of-the-art method.

59 BASIC BIOLOGICAL SCIENCES↗

In situ Detection of Plasma Induced Surface Interaction based on Deep Learning based Visual Diagnostics (Technical Report)

It is characteristic for many plasma devices to undergo plasma-material interaction leading to surface erosion. These processes, often not easily detectable, lead to changes in device performance and lifespan. State-of-the-art lifetime tests and wear experiments require over 1000s hours. A self-consistent model for accurately predicting the erosion's effects is not available. In situ detection of these processes is not a trivial task since the surface variations at the early stages have a micron scale. Such limitations not only restrict testing and prediction capabilities but also slow the development of new thrusters and limit mission duration. To address these challenges, an in-situ diagnostic for real-time erosion assessment has been developed, aiming to expedite lifetime testing and broaden experimental campaigns. Several works were dedicated to real-time and in situ monitoring of material erosion during plasma exposure using laser holography, microscopy, and with telemicroscopes. However, the applicability of these approaches is limited due to complexity, cost and less flexibility as they often require placing diagnostic equipment inside the vacuum chamber. In collaboration with Princeton Collaborative Research Facility (PCRF), Princeton Plasma Physics Laboratory (PPPL), a new diagnostic approach is developed, where geometry modifications to the ceramic channel walls were introduced that would result in accelerated channel erosion. We employed Long-distance microscope (LDM) imagery, combined with Deep-Learning based Shape from focus or depth from focus (DFF or SFF) approach, that provides an accessible and cost-effective solution. LDM employs focus variation techniques to continuously capture multiple images of the target object at distinct focal planes. DFF, an optical focus variation method, generates a 3D topographical surface depth map from a sequence of variably focused images. Combined with the developed diagnostic, this approach offers a controllable means to study erosion under accelerated conditions. In this work, we develop Neural Network-based DFF algorithm applicable for LDM data to quantitatively evaluate plasma induced surface modification from LDM data. Next, we develop Deep Learning-based super-resolution depth map image reconstruction technique to increase the resolution of depth maps obtained from DFF algorithm to improve the accuracy of erosion measurements. Thirdly, we develop several image processing techniques to remove noise and improve the quality of depth map image. Here we report the results of initial tests for this approach. An experimental setup designed and built in PPPL was employed that consists of a 3-cm gridded ion source that produces a neutralized argon beam with energies up to 600 eV. A hexagonal boron nitride (h-BN) ceramic target, designed based on computational predictions, was used. Tests were conducted to reconstruct the complex geometry of the target under the lighting conditions of the operated ion source.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

The Twins Embedding of Type Ia Supernovae. II. Improving Cosmological Distance Estimates

We show how spectra of Type Ia supernovae (SNe Ia) at maximum light can be used to improve cosmological distance estimates. In a companion article, we used manifold learning to build a three-dimensional parameterization of the intrinsic diversity of SNe Ia at maximum light that we call the "Twins Embedding."In this article, we discuss how the Twins Embedding can be used to improve the standardization of SNe Ia. With a single spectrophotometrically calibrated spectrum near maximum light, we can standardize our sample of SNe Ia with an rms of 0.101 0.007 mag, which corresponds to 0.084 0.009 mag if peculiar velocity contributions are removed and to 0.073 0.008 mag if a larger reference sample were obtained. Our techniques can standardize the full range of SNe Ia, including those typically labeled as peculiar and often rejected from other analyses. We find that traditional light-curve width + color standardization such as SALT2 is not sufficient. The Twins Embedding identifies a subset of SNe Ia, including, but not limited to, 91T-like SNe Ia whose SALT2 distance estimates are biased by 0.229 0.045 mag. Standardization using the Twins Embedding also significantly decreases host-galaxy correlations. We recover a host mass step of 0.040 0.020 mag compared to 0.092 0.026 mag for SALT2 standardization on the same sample of SNe Ia. These biases in traditional standardization methods could significantly impact future cosmology analyses if not properly taken into account.

79 ASTRONOMY AND ASTROPHYSICS↗

Cohort organized learning: clustering through agreement

In this article we describe cohort organized learning (CoOL), a method for clustering data without explicit distance or similarity computations. Herein, we will describe CoOL, derive the gradients determined by expectation maximization to train the networks, show how to monitor convergence during training and evaluate the clusters after training, and discuss a series of examples and use cases. We also discuss CoOL’s limitations and future prospects on related tasks. Because CoOL uses neural networks to estimate the clusters, it can be used to cluster any data that can be made compatible and we illustrate this on vector data and images.

clustering↗

Nearest-Neighbor Machine Learning Feature Selection for Interpretation of Microbial Molecular Signatures from Isotope Ratio Mass Spectrometry Data

Mass spectrometry (MS) promises to be a powerful tool for potential biosignature detection during astrobiological missions on ocean worlds in our solar system. Accurate and generalizable machine learning methods could enhance science return on investment by predicting seawater chemistry and classifying isotopic biosignatures, either as a signature consistent with microbial life (biotic) or as a novelty (unclassified/unique). However, machine learning models are likely to be complex and involve interactions between MS features, making biosignatures difficult to interpret. Feature selection methods provide biological and chemical context that help interpret the mechanisms of machine learning models, but these methods also need the ability to detect complex interactions. Previously, we developed a machine learning feature selection algorithm called nearest-neighbor projected distance regression (NPDR) that has the ability to identify important model features that involve complex interactions and automatically reduce correlation and the dimensionality in a high-dimensional variable space. The standard distance metrics used in NPDR – Manhattan and Euclidean – assume the multivariate data are isotropic, which is often violated in real data due to differences in the covariance between variables. Thus, we extend NPDR to include a random forest distance, and other anisotropic distance metrics, for computing nearest neighbors. We also augment the isotope-ratio MS data with time-series features from the raw MS signal to improve biotic classification. We test NPDR on our novel experimental ocean world seawater analog MS data. We measure isotope fractionations of volatile CO 2 that could be measured in exospheres or plumes. Samples include baseline abiotic conditions using a range of possible seawater chemistry consistent with Europa and Enceladus, and biotic samples that include microbes in these seawaters. We use penalized NPDR with random forest proximity to identify interpretable microbial molecular signatures. We compare features with random forest importance, and we train a classifier that discriminates between biotic and abiotic samples with high accuracy. These ML-trained ocean-world analog MS data could be used to assist in identifying biosignatures during future missions.

geochemistry↗

Quantum-assisted associative adversarial network: applying quantum annealing in deep learning

Abstract Generative models have the capacity to model and generate new examples from a dataset and have an increasingly diverse set of applications driven by commercial and academic interest. In this work, we present an algorithm for learning a latent variable generative model via generative adversarial learning where the canonical uniform noise input is replaced by samples from a graphical model. This graphical model is learned by a Boltzmann machine which learns low-dimensional feature representation of data extracted by the discriminator. A quantum processor can be used to sample from the model to train the Boltzmann machine. This novel hybrid quantum-classical algorithm joins a growing family of algorithms that use a quantum processor sampling subroutine in deep learning, and provides a scalable framework to test the advantages of quantum-assisted learning. For the latent space model, fully connected, symmetric bipartite and Chimera graph topologies are compared on a reduced stochastically binarized MNIST dataset, for both classical and quantum sampling methods. The quantum-assisted associative adversarial network successfully learns a generative model of the MNIST dataset for all topologies. Evaluated using the Fréchet inception distance and inception score, the quantum and classical versions of the algorithm are found to have equivalent performance for learning an implicit generative model of the MNIST dataset. Classical sampling is used to demonstrate the algorithm on the LSUN bedrooms dataset, indicating scalability to larger and color datasets. Though the quantum processor used here is a quantum annealer, the algorithm is general enough such that any quantum processor, such as gate model quantum computers, may be substituted as a sampler.

Wilson, Max (ORCID:0000000207983391)↗

SU-D-BRB-01: A Comparison of Learning Methods for Knowledge Based Dose Prediction for Coplanar and Non-Coplanar Liver Radiotherapy

Purpose: The predictive power of knowledge based planning (KBP) has considerable potential in the development of automated treatment planning. Here, we examine the predictive capabilities and accuracy of previously reported KBP methods, as well as an artificial neural networks (ANN) method. Furthermore, we compare the predictive accuracy of these methods on coplanar volumetric-modulated arc therapy (VMAT) and non-coplanar 4π radiotherapy. Methods: 30 liver SBRT patients previously treated using coplanar VMAT were selected for this study. The patients were re-planned using 4π radiotherapy, which involves 20 optimally selected non-coplanar IMRT fields. ANNs were used to incorporate enhanced geometric information including liver and PTV size, prescription dose, patient girth, and proximity to beams. The performance of ANN was compared to three methods from statistical voxel dose learning (SVDL), wherein the doses of voxels sharing the same distance to the PTV are approximated by either taking the median of the distribution, non-parametric fitting, or skew-normal fitting. These three methods were shown to be capable of predicting DVH, but only median approximation can predict 3D dose. Prediction methods were tested using leave-one-out cross-validation tests and evaluated using residual sum of squares (RSS) for DVH and 3D dose predictions. Results: DVH prediction using non-parametric fitting had the lowest average RSS with 0.1176(4π) and 0.1633(VMAT), compared to 0.4879(4π) and 1.8744(VMAT) RSS for ANN. 3D dose prediction with median approximation had lower RSS with 12.02(4π) and 29.22(VMAT), compared to 27.95(4π) and 130.9(VMAT) for ANN. Conclusion: Paradoxically, although the ANNs included geometric features in addition to the distances to the PTV, it did not perform better in predicting DVH or 3D dose compared to simpler, faster methods based on the distances alone. The study further confirms that the prediction of 4π non-coplanar plans were more accurate than VMAT. NIH R43CA183390 and R01CA188300.

60 APPLIED LIFE SCIENCES↗

Final Report for Award DE-SC0005403. Improved Electrochemical Performance of Strained Lattice Electrolytes via Modulated Doping

The enclosed document provides a final report to document the research performed at the University of Delaware under Grant DE-SC0005403: Improved Electrochemical Performance of Strained Lattice Electrolytes via Modulated Doping. The ultimate goal of this project was to learn how to systematically strain the inter-atomic distance in thin ceramic films and how to use this newfound control to improve the ease by which oxygen ions can conduct through the films. Increasing the ionic conductivity of ceramics holds the promise of drastic improvements in the performance of solid oxide fuel cells, chemical sensors, gas permeation membranes, and related devices. Before this work, the experimental evidence advocating for strain-based techniques was often controversial and poorly characterized. Enabling much of this work was a new method to quickly create a very wide range of ceramic nanostructures that was established during the first phase of the project. Following this initial phase, we created a variety of promising nanostructured epitaxial films and multilayers with systematic variations in lattice mismatch and dopant content. Over the course of the work, a positive effect of tensile atomic strain on the oxygen conductivity was conclusively found using a few different forms of samples and experimental techniques. The samples were built by sputtering, an industrially scalable technique, and thus the technological implementation of these results may be economically feasible. Still, two other results consistently achieved over multiple efforts in this work give pause. The first of these results was that very specific, pristine surfaces upon which to build the nanostructures were strictly required in order to achieve measurable results. The second of these results was that compressively strained films with concomitant reductions in oxygen conductivity are much easier to obtain relative to tensile-strained films with increased conductivity.

36 MATERIALS SCIENCE↗

Vehicular Re-Identification from Uncontrolled Multiple Views

Vehicle re-identification (re-ID) across disparate sensing modalities remains a fundamental challenge for transportation research. In this work, we introduce a deep multi-view vehicle re-ID framework that leverages Siamese networks to compare pairs of vehicle images and produce matching scores, enabling robust association across drastically different viewpoints such as those from UAVs, surveillance cameras, and ground sensors. The model exploits convolutional neural networks to learn features that remain discriminative under changes in angle, distance, and illumination, supporting more generalizable re-ID performance. As part of this effort, we also developed an automated pipeline to synchronize roadside and UAV video streams, producing a multi-perspective dataset that complements preexisting real collections and a synthetic dataset generated in this study. Together, these contributions advance the capability to re-identify vehicles across wide viewing baselines; establish a foundation for scalable, reproducible research in vehicle re-ID; and open pathways for future applications, such as inferring routine behaviors, movement patterns, and daily habits of the individual associated with the vehicle.

convolutional neural networks↗