Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Fast Timing”

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 1,207 records · Page 67
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Simulation of Li-pellet triggered ELMs in EAST with an impurity model implemented under BOUT++ framework

A simple impurity model has been developed under the BOUT++ framework to investigate the Li-pellet triggered edge localized mode (ELM) in EAST configuration. The present impurity model decouples the ion pressure enhancement effect (IPEE) (which affects radial E X B flow shear and gyro-viscosity of deuterium ions and is implicitly included in our previous work) to make the simulation of Li pellet injection more physically reasonable. In addition, the present impurity model also includes the impurity equilibrium effect (IEE) (which induces modifications on vorticity and gyro-viscosity). The simulation results show that without IPEE a turbulent ELM induced by Li pellet is triggered by multiple peeling-ballooning modes (PBMs) rather than by a single dominant mode, and the simultaneous growth of multiple modes is conducive to reducing the time for pedestal entering the energy loss state. When a turbulent ELM occurs, the nonlinearly dominant modes undergo a secondary fast growth during fast crash phase. The finding explains the secondary increase of D α emission observed in the DIII-D Li injection experiments. During the evolution of a turbulent ELM, the continuous pedestal collapse dominates the pedestal energy loss in turbulent transport phase; however, in saturation phase, the stabilizing effects on PBMs (n ≠ 0) by n = 0 mode and radial E X B flow shear within the steepest gradient region are more prominent. It is also found that with IEE and without IPEE, the size threshold of ELM triggering and the magnitude of ELM sizes both show a good agreement with EAST and DIII-D experimental observations.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2025

Investigating performance and variability of NIF ICF experiments with deep learning

The parameter space involved in designing an inertial confinement fusion shot at the National Ignition Facility (NIF) is massively multi-dimensional and the cost of a single shot makes a comprehensive set of sensitivity studies in the laboratory impractical. The use of machine learning to overcome these challenges has gained popularity and has had several successful applications by the scientific community. We extend on these efforts by training a neural network (NN) on information about the experimental design, engineering elements, and drive asymmetry to predict with uncertainty the neutron yield of an experiment. We find the measured and model predicted values are in good agreement, with an R 2 value of 0.91 for a randomly selected test dataset. Almost all the predicted 95% credible intervals contain the corresponding measured value for both training and test datasets. We identify correlations picked up by the NN between the shot design, yield, and variability and use them to motivate shot sensitivity studies. The first shot to exceed the Lawson-like ignition criteria (N210808) was conducted at the NIF and subsequent shots studied the design’s robustness. In a follow-up shot to N210808, our model predicts capsule quality to be the main performance degradation mechanism that prevented the shot from repeating previous performance levels. Shot N221204 was the first shot to exceed a target energy gain of 1. Our model predicts increased yield with reduced coast time for a N221204 study and greater variability for designs with lower peak powers at constant yield. The model’s fast prediction speed and uncertainty prediction are useful for identifying interesting design paths that could warrant further investigation with conventional simulations to search for robust high yield designs.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2021

Measurement of proton quenching in a plastic scintillator detector

The non-linear energy response of the plastic scintillator EJ-260 is measured with the MicroCHANDLER detector, using neutron beams of energy 5 to 27 MeV at the Triangle Universities Nuclear Laboratory. The first and second order Birks' constants are extracted from the data, and found to be $k_B = (8.70 \pm 0.93)\times 10^{-3}\ {\rm g/cm^2/MeV}$ and $k_C = (1.42 \pm 1.00) \times 10^{-5}\ {\rm (g/cm^2/MeV)^2}$. This result covers a unique energy range that is of direct relevance for fast neutron backgrounds in reactor inverse beta decay detectors. Furthermore, these measurements will improve the energy non-linearity modeling of plastic scintillator detectors. In particular, the updated energy response model will lead to an improvement of fast neutron modeling for detectors based on the CHANDLER reactor neutrino detector technology.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

Multiparticle cumulant mapping for Coulomb explosion imaging: Calculations and algorithm

We present a versatile cumulant mapping algorithm for analyzing correlated particle emission, offering insights into complex electronic and nuclear dynamics. Recently, we have demonstrated the use of cumulant mapping to extract information-rich correlations between the momenta of multiple fragments produced in Coulomb explosion imaging experiments [C. Cheng et al., Phys. Rev. Lett. 130, 093001 (2023)]. We define cumulant mapping in terms of histograms, enabling fast computation of linear (additive) observables. However, applying the same algorithm to nonlinear (nonadditive) observables poses challenges, as the computation time of conventional estimators scales nonlinearly with data size. To overcome this, we develop estimators and an accompanying algorithm to enable computationally efficient estimation of the cumulant of interest. Comparisons of computation times and signal-to-noise ratios reveal the superior performance of our approach. This method is demonstrated on the (D+, D+, C+, O+) dissociation channel of CD 2 ⁢O 4+ produced in a strong-field ionization experiment. Additionally, Poisson statistics are used to simulate the two methods and provide insights into the efficiency of our algorithm. The proposed methodology unlocks efficient computation of cumulant mapping for a broader range of complex systems and observables, such as the laser pulse dependence of ionization dynamics.

74 ATOMIC AND MOLECULAR PHYSICS↗
DOE OSTI2020

Fast nonthermal processes in pulsed laser deposition

Pulsed laser deposition (PLD) is typically used to grow epitaxial thin films of quantum materials. In this research, we use in situ x-ray scattering to study homoepitaxy of SrTiO 3 by energetic deposition (e-PLD) versus PLD thermalized by a He background gas (th-PLD). Energetic PLD suppresses the lateral growth of two-dimensional islands, which suggests that particles with kinetic energies of ~ 100 eV break up smaller islands. Fast interlayer transport occurs for th-PLD as well as e-PLD, implying a process operating on submicrosecond time scales for incident particles with kinetic energies below 10 eV.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Laser-driven ion and electron acceleration from near-critical density gas targets: Towards high-repetition rate operation in the 1 PW, sub-100 fs laser interaction regime

Ion acceleration from gaseous targets driven by relativistic-intensity lasers was demonstrated as early as the late 1990s, yet most of the experiments conducted to date have involved picosecond-duration, Nd:glass lasers operating at low repetition rate. Here, we present measurements on the interaction of ultraintense ( ≈ 10 20 W cm − 2 , 1 PW ) , ultrashort ( ≈ 70 fs ) Ti:Sa laser pulses with near-critical ( ≈ 10 20 cm − 3 ) helium gas jets, a debris-free targetry with the potential for future compatibility with high ( ≈ 1 Hz ) repetition rate operation. We provide evidence of α particles being forward accelerated up to ≈ 2.7 − MeV energy with a total flux of ≈ 10 11 sr − 1 as integrated over > 0.1 − MeV energies and detected within a 0.5 − mrad solid angle. We also report on on-axis emission of relativistic electrons with an exponentially decaying spectrum characterized by a ≈ 10 − MeV slope, i.e., five times larger than the standard ponderomotive scaling. The total charge of these electrons with energy above 2 MeV is estimated to be of ≈ 1 nC , corresponding to ≈ 0.1 % of the laser drive energy. In addition, we observe the formation of a plasma channel, extending longitudinally across the gas density maximum and expanding radially with time. These results are well captured by large-scale particle-in-cell simulations, which reveal that the detected fast ions most likely originate from reflection off the rapidly expanding channel walls. The latter process is predicted to yield ion energies in the MeV range, which compare well with the measurements. Finally, direct laser acceleration is shown to be the dominant mechanism behind the observed electron energization. Published by the American Physical Society 2024

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2022

Improving Schrödinger Equation Implementations with Gray Code for Adiabatic Quantum Computers

We reformulate the continuous-space Schrödinger equation in terms of spin Hamiltonians. For the kinetic energy operator, the critical concept facilitating the reduction in model complexity is the idea of position encoding. A binary encoding of position produces a spin-1/2 Heisenberg-like model and yields exponential improvement in space complexity when compared to classical computing. Encoding with a binary reflected Gray code (BRGC), and a Hamming-distance-2 Gray code (H2GC) reduces the model complexity down to the 𝑋⁢𝑍 and transverse Ising model, respectively. For 𝐴 qubits BRGC yields 2 𝐴 positions and is reduced to its 2-local form with O⁡(𝐴) ancillary qubits. H2GC yields 2 𝐴/2+1 positions with O⁡(𝐴 2 ) three-local penalty terms. We also identify the bijective mapping between diagonal unitaries and the Walsh series, producing the mapping of any real potential to a series of 𝑘 -local Ising models through the fast Walsh transform. Finally, in a finite volume, we provide some numerical evidence to support the claim that the total time needed for adiabatic evolution is protected by the infrared cutoff of the system. As a result, initial state preparation from a free-field wave function to an interacting system is expected to exhibit polynomial time complexity with volume and constant scaling with respect to lattice discretization for all encodings. For H2GC, if the evolution starts with the transverse Hamiltonian due to hardware restrictions, then penalties are dynamically introduced such that the low-lying spectrum reproduces the energy levels of the Laplacian. The adiabatic evolution of the penalty Hamiltonian is therefore sensitive to the ultraviolet scale. It is expected to exhibit polynomial time complexity with lattice discretization, or exponential time complexity with respect to the number of qubits given a fixed volume.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

2 kV, 0.7 mΩ•cm2 Vertical Ga2O3 Superjunction Schottky Rectifier with Dynamic Robustness

We report the first experimental demonstration of a vertical superjunction device in ultra-wide bandgap (UWBG) Ga 2 O 3 . The device features 1.8 μm wide, 2×10 17 cm -3 doped n-Ga 2 O 3 pillars wrapped by the charge-balanced p-type nickel oxide (NiO). The sidewall NiO is sputtered through a novel self-align process. Benefited from the high doping in Ga 2 O 3 , the superjunction Schottky barrier diode (SJ-SBD) achieves a ultra-low specific on-resistance (R ON,SP ) of 0.7 mΩ•cm 2 with a low turn-on voltage of 1 V and high breakdown voltage (BV) of 2000 V. The R ON,SP ~BV trade-off is among the best in all WBG and UWBG power SBDs. The device also shows good thermal stability with BV > 1.8 kV at 175 °C. In the unclamped inductive switching tests, the device shows a dynamic BV of 2.2 kV and no degradation under 1.7 kV repetitive switching, verifying the fast acceptor depletion in NiO under dynamic switching. Such high-temperature and switching robustness are reported for the first time in a heterogeneous superjunction. These results show the great potential of UWBG superjunction power devices.

Qin, Yuan↗
DOE OSTI2020

Destructive PIE and Safety Testing of Six AGR-2 UO 2 Capsule 3 Compacts

The Advanced Gas Reactor (AGR) Fuel Development and Qualification Program’s second irradiation experiment (AGR-2) was irradiated in the Advanced Test Reactor (ATR) at Idaho National Laboratory (INL) from June 2010 to October 2013 (Collin 2014). The fuel compacts in this experiment held either tristructural isotropic (TRISO)-coated spherical kernels of uranium oxide (UO2) or TRISO-coated kernels containing both uranium carbide and uranium oxide phases (UCO). There were six separately monitored and controlled capsules in the AGR-2 test train. Capsule 3 held twelve compacts containing UO2-TRISO particles fabricated by BWX Technologies Nuclear Operations Group. The AGR-2 TRISO particles were fabricated in a pilot-scale fluidized-bed chemical vapor deposition (FB-CVD) furnace with a coating chamber inner diameter of 150 mm (Phillips, Barnes, and Hunn 2010), which was a change from the first irradiation experiment (AGR-1) particles that had been coated in a lab-scale FB-CVD coating system with a chamber inner diameter of 50 mm (Lowden 2006). The TRISO particles were overcoated with resinated graphite flake at Oak Ridge National Laboratory (ORNL), and the overcoated particles were pressed into one-inch-long, half-inch-diameter cylinders (Hunn, Montgomery, and Pappano 2010). Each cylindrical compact held an average of 1,543 TRISO particles with 9.6% enriched UO2 kernels that had a nominal diameter of 500 μm (Hunn, Savage, and Silva 2012). Capsule 3 compacts were irradiated to average calculated burnups of 9.01–10.69% fissions per initial metal atom (FIMA), and the average calculated fluences of fast neutrons with energies E n > 0.18 MeV were 3.05–3.53×10 25 n/m 2 (Sterbentz 2014). The calculated time-average, volume-average Capsule 3 compact temperatures were 996–1,062°C. However, Capsule 3 compact temperatures varied several hundred degrees across each compact, and the timeaverage minimum (TA min ) and time-average maximum (TA max ) temperatures were between 889–999°C and 1,072–1,105°C, respectively (Hawkes 2014). After irradiation, the AGR-2 test train was transferred from ATR to the INL Materials and Fuels Complex for inspection and disassembly (Ploger, Demkowicz, and Harp 2015). The initial inspection included dimensional metrology of the compacts and graphite fuel holders. Like all the AGR-2 compacts, the compacts in Capsule 3 shrank slightly during irradiation, as expected, with an average length reduction of 1.07–1.24% and an average diameter reduction of 0.13–0.41%. Post-irradiation examination (PIE) of the capsule components was completed to measure select fission products ( 90 Sr, 110m Ag, 134 Cs, 137 Cs, 144 Ce, and 154 Eu) outside the compacts (Stempien and Demkowicz 2020). This involved gamma counting of the graphite and graphoil spacers at the top and bottom of each capsule, acid leaching for radiochemical analysis of fission products on the metallic capsule components, and burn-leach analysis of the graphite holders. The total amount of 110mAg measured on the Capsule 3 components was 13% of the calculated capsule inventory. This was significantly lower than the amount of 110m Ag measured on the three UCO capsule components, which ranged from 32–70%. The lower 110m Ag release in Capsule 3 was likely due to lower peak temperatures compared with the UCO fuel capsules (Hawkes 2014). Measured inventories of the other select fission products were also lower in Capsule 3.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2024

Refining Fast Simulation Using Machine Learning

A growing reliance on the fast Monte Carlo (FastSim) will accompany the high luminosity and detector granularity expected in Phase 2. FastSim is roughly 10 times faster than equivalent GEANT4-based full simulation (FullSim). However, reduced accuracy of the FastSim affects some analysis variables and collections. To improve its accuracy, FastSim is refined using regression-based neural networks trained with ML. The status of FastSim refinement is presented. The results show improved agreement with the FullSim output and an improvement in correlations among output observables and external parameters.

Güngördü, Acelya Deniz↗
DOE OSTI2020

JWST Noise Floor. I. Random Error Sources in JWST NIRCam Time Series

James Webb Space Telescope (JWST) transmission and emission spectra will provide invaluable glimpses of transiting exoplanet atmospheres, including possible biosignatures. This promising science from JWST, however, will require exquisite precision and understanding of systematic errors that can impact the time series of planets crossing in front of and behind their host stars. Here, we provide estimates of the random noise sources affecting JWST Near-Infrared Camera (NIRCam) time-series data on the integration-to-integration level. We find that 1/f noise can limit the precision of grism time series for two groups (230–1000 ppm depending on the extraction method and extraction parameters), but will average down like the square root of N frames/reads. The current NIRCam grism time-series mode is especially affected by 1/f noise because its GRISMR dispersion direction is parallel to the detector fast-read direction, but could be alleviated in the GRISMC direction. Care should be taken to include as many frames as possible per visit to reduce this 1/f noise source: thus, we recommend the smallest detector subarray sizes one can tolerate, four output channels, and readout modes that minimize the number of skipped frames (RAPID or BRIGHT2). We also describe a covariance-weighting scheme that can significantly lower the contributions from 1/f noise as compared to sum extraction. We evaluate the noise introduced by pre-amplifier offsets, random telegraph noise, and high dark current resistor capacitor (RC) pixels and find that these are correctable below 10 ppm once background subtraction and pixel masking are performed. We explore systematic error sources in a companion paper.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2021

Low-Cost Transformation of Biomass-Derived Carbon to High-Performing Nano-graphite via Low-Temperature Electrochemical Graphitization

Graphite, an essential component of energy storage devices, is traditionally synthesized via an energy-intensive thermal process (Acheson process) at ~3300 K. However, the battery performance of such graphite is abysmal under fast-charging conditions, which is deemed essential for the propulsion of electric vehicles to the next level. Herein, a low-temperature electrochemical transformation approach has been demonstrated to afford a highly crystalline nano-graphite with the capability of tuning interlayer spacing to enhance the lithium diffusion kinetics in molten salts at 850 °C. The essence of our strategy lies in the effective electrocatalytic transformation of carbon to graphite at a lower temperature that could significantly increase the energy savings, reduce the cost, shorten the synthesis time, and replace the traditional graphite synthesis. The resulting graphite exhibits high purity, crystallinity, a high degree of graphitization, and a nanoflake architecture that all ensure fast lithium diffusion kinetics (~2.0 × 10 –8 cm2 s –1 ) through its nanosheet. Such unique features enable outstanding electrochemical performance (~200 mA h g –1 at 5C for 1000 cycles, 1C = 372 mA g –1 ) as a fast-charging anode for lithium-ion batteries. Overall, this finding paves the way to make high energy-density fast-charging batteries that could boost electromobility.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Optically multiplexed neutron time-of-flight technique for inertial confinement fusion

Neutron time-of-flight (nTOF) detectors are crucial in diagnosing the performance of inertial confinement fusion (ICF) experiments, which implode targets of deuterium–tritium fuel to achieve thermonuclear conditions. These detectors utilize the fusion neutron energy spectrum to extract key measurements, including the hotspot ion temperature and fuel areal density. Previous work [Danly et al., Rev. Sci. Instrum. 94, 043502 (2023)] has demonstrated adding 1D spatial resolution to an nTOF-like detector using a neutron aperture and streak camera to measure the ion temperature profile of an ICF implosion. By contrast, the study presented herein explores modifying the 1D detector to use a fast photomultiplier tube (PMT) to validate the design of a 2D spatially resolved instrument based on reconstruction from 1D profiles. The modification would collect time-of-flight traces from separate scintillators in an imaging array with one PMT using optical fibers of varying lengths to time-multiplex the signals. This technique has been demonstrated in ride-along experiments on the OMEGA laser with 20 fiber-coupled scintillator channels connected to a Photek PMT210. Results provide constraints on the fiber lengths and PMT gating requirements to promote pulse fidelity throughout all channels. Calibration of the detector to fixed nTOFs can provide a preliminary estimate of the instrument response function (IRF), although measurement of the IRF is currently under way. These results suggest that nTOF signals can potentially be time-multiplexed with fibers so long as the design is strategic to mitigate signal-to-noise reduction, modal dispersion, and charge build-up in the PMT, which has implications beyond ion temperature imaging.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2025

Sub-microsecond Transformers for Jet Tagging on FPGAs

We present the first sub-microsecond transformer implementation on an FPGA achieving competitive performance for state-of-the-art high-energy physics benchmarks. Transformers have shown exceptional performance on multiple tasks in modern machine learning applications, including jet tagging at the CERN Large Hadron Collider (LHC). However, their computational complexity prohibits use in real-time applications, such as the hardware trigger system of the collider experiments up until now. In this work, we demonstrate the first application of transformers for jet tagging on FPGAs, achieving $\mathcal{O}(100)$ nanosecond latency with superior performance compared to alternative baseline models. We leverage high-granularity quantization and distributed arithmetic optimization to fit the entire transformer model on a single FPGA, achieving the required throughput and latency. Furthermore, we add multi-head attention and linear attention support to hls4ml, making our work accessible to the broader fast machine learning community. This work advances the next-generation trigger systems for the High Luminosity LHC, enabling the use of transformers for real-time applications in high-energy physics and beyond.

Laatu, Lauri [Imperial Coll., London]↗
DOE OSTI2023

Effect of Si Content on Extreme Fast Charging Behavior in Silicon–Graphite Composite Anodes

Commercial Li-ion batteries typically incorporate a small amount of high-capacity silicon (Si)-based materials in the composite graphite-based anode to increase the energy density of the battery. However, very little is known about the effects of Si on the fast-charging behavior of composite anodes. Herein, we examine the effects of the Si/graphite ratio in the composite anode on the fast-charging behavior of full cells. We show that addition of Si increases the rate capability from 1C to 8C and improves the capacity retention in early cycles at 6C due to reduced overpotential in constant current charging cycles. The impacts of Si content on fast-charging aging were identified by Post-Test characterization. Despite realizing benefits of available capacity and reduced Li plating at 6C, silicon–electrolyte interactions lead the time-dependent cell performance to fade quickly in the long term. The Post-Test analysis also revealed the thickening of the electrode and nonuniform distribution of electrolyte decomposition products on the Si-containing anodes, as well as the organic-rich solid electrolyte interphase (SEI), which are the factors behind cell degradation. Our study sheds insight on the advantages and disadvantages of Si/graphite composite anodes when they are used in fast-charging applications and guides further research in the area by designing an optimized composition of Si incorporated in a mature graphite matrix.

25 ENERGY STORAGE↗
DOE OSTI2025

A Particle-in-Cell Method for Plasmas with a Generalized Momentum Formulation, Part I: Model Formulation

Here, this paper formulates a new particle-in-cell method for the Vlasov–Maxwell system. Under the Lorenz gauge condition, Maxwell’s equations for the electromagnetic fields can be written as a collection of scalar and vector wave equations. The use of potentials for the fields motivates the adoption of a Hamiltonian formulation for particles that employs the generalized (conjugate) momentum. A notable advantage offered by the Hamiltonian formulation is the elimination of time derivatives in the Lorenz gauge formulation that are required by the standard Newton–Lorentz treatment of the particles. This allows the fields to retain the full time-accuracy guaranteed by the field solver. The resulting updates for particles require only knowledge of the fields and their spatial derivatives. An analytical method for constructing these spatial derivatives is presented that exploits the underlying integral solution used in the field solver for the wave equations. Moreover, these derivatives are demonstrated to converge at the same rate as the fields in both time and space. The Method of Lines Transpose field solver we consider in this work is globally first-order accurate in time and high-order accurate in space (e.g., fourth- and fifth-order) and belongs to a larger class of methods which are unconditionally stable, can address geometry, and leverage $\mathcal {O}(N)$ fast summation methods for efficiency. We demonstrate the method on several well-established benchmark problems on bounded domains, including a plasma sheath as well as a relativistic particle beam. The efficacy of the proposed formulation is established by comparing with a second-order accurate finite-difference time-domain method that employs a leapfrog time advance for particles and a charge conserving map suitable for bounded domains. The new method shows mesh-independent numerical heating properties even in cases where the plasma Debye length is smaller than the grid spacing. This is an important feature of the new method for problems defined on bounded domains, because it permits the use of coarser grids in space in the representation of the fields. Such a capability has significant implications for the simulation of plasmas in bounded domains with complex geometry, where the ratio between the largest and smallest cells can vary significantly. The use of high-order spatial approximations in the new method also means that fewer grid points are required in order to achieve a fixed accuracy. Our results also suggest that the new method can be used with fewer simulation particles per cell compared to the benchmark explicit method, which permits further computational savings.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Grid Architecture Guidance Specification for FAST-DERMS (V.1.0)

This document provides grid architecture guidance to support the Grid Modernization Laboratory Consortium (GMLC) Federated Architecture for Secure and Transactive Distributed Energy Resource Management Solution (FAST-DERMS) project. The FAST-DERMS project seeks to develop an architecture and reference implementation solution that enables the provision of reliable, resilient, and secure transmission and distribution (T&D) grid services through the scalable aggregation and near-real-time management of utility-scale and small-scale distributed energy resources (DERs).

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2025

Active Learning Meets Foundation Models: Fast Remote Sensing Data Annotation for Object Detection

Object detection in remote sensing demands extensive, high-quality annotations—a process that is both labor-intensive and time-consuming. In this work, we introduce a real-time active learning and semi-automated labeling framework that leverages foundation models to streamline dataset annotation for object detection in remote sensing imagery. For example, by integrating a Segment Anything Model (SAM), our approach generates mask-based bounding boxes that serve as the basis for dual sampling: (a) uncertainty estimation to pinpoint challenging samples, and (b) diversity assessment to ensure broad data coverage. Furthermore, our Dynamic Box Switching Module (DBS) addresses the well-known cold start problem for object detection models by replacing its suboptimal initial predictions with SAM-derived masks, thereby enhancing early-stage localization accuracy. Extensive evaluations on multiple remote sensing datasets plus a real-world user study, demonstrate that our framework not only reduces annotation effort, but also significantly boosts detection performance compared to traditional active learning sampling methods. The code for training and the user interface will be made available.

Burges, Marvin [ORNL] (ORCID:0000000312690769)↗