Search NASA⌕ Search

SEARCH · Search NASA

Results for “compressed training”

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 73 records · Page 4
All sourcesNASA NTRSDOE OSTI
NASA NTRS2010

Speech Acquisition and Automatic Speech Recognition for Integrated Spacesuit Audio Systems

A voice-command human-machine interface system has been developed for spacesuit extravehicular activity (EVA) missions. A multichannel acoustic signal processing method has been created for distant speech acquisition in noisy and reverberant environments. This technology reduces noise by exploiting differences in the statistical nature of signal (i.e., speech) and noise that exists in the spatial and temporal domains. As a result, the automatic speech recognition (ASR) accuracy can be improved to the level at which crewmembers would find the speech interface useful. The developed speech human/machine interface will enable both crewmember usability and operational efficiency. It can enjoy a fast rate of data/text entry, small overall size, and can be lightweight. In addition, this design will free the hands and eyes of a suited crewmember. The system components and steps include beam forming/multi-channel noise reduction, single-channel noise reduction, speech feature extraction, feature transformation and normalization, feature compression, model adaption, ASR HMM (Hidden Markov Model) training, and ASR decoding. A state-of-the-art phoneme recognizer can obtain an accuracy rate of 65 percent when the training and testing data are free of noise. When it is used in spacesuits, the rate drops to about 33 percent. With the developed microphone array speech-processing technologies, the performance is improved and the phoneme recognition accuracy rate rises to 44 percent. The recognizer can be further improved by combining the microphone array and HMM model adaptation techniques and using speech samples collected from inside spacesuits. In addition, arithmetic complexity models for the major HMMbased ASR components were developed. They can help real-time ASR system designers select proper tasks when in the face of constraints in computational resources.

Huang, Yiteng↗
NASA NTRSDate not supplied

Isolator Shock Train Response to Dual-Mode Scramjet Throttling-Like Forcing

Typical dual-mode scramjet designs employ an isolator component situated between the inlet and combustor. The isolator fluidically buffers the inlet from the combustor via the formation of a shock train, an interconnected series of shock-boundary layer interactions. The isolator is required during ramjet-engine operation and at the lower edges of the scramjet operational envelope. In an operational engine, transient behavior imposes fluidic dynamics due to attitude adjustments, combustor throttling, aberrant resonances within the combustor, etc. To accommodate these transients, the shock train adjusts on both temporal and spatial scales. Sufficiently large downstream pressure increases in the combustor play a critical role during unstart, whereby strong compression waves propagate upstream in the isolator to ultimately force the shock train out the inlet, if it cannot stabilize to the pressure increase. This critically reduces the engine mass flow rate, potentially causing a catastrophic reduction in thrust and loss of the vehicle. A nearly instantaneous unstart can, in principle, be predicted by a ratio of the normal shock pressure jump in an inviscid isolator. However, lower-magnitude perturbations, below this bound, also induce unstarts. Understanding the transient behavior inherent in unsteady isolator operation is critical to reducing safety margin in isolator length and refining engine control algorithms to improve engine performance. The Isolator Dynamics Research Lab (IDRL) is a cold-flow, direct-connect isolator facility ideal for assessing the fundamental fluid dynamics of shock trains. The internal flow of a scramjet isolator is modeled using a 2D, symmetric Mach 2.5 nozzle followed by a 695 mm long, 25.4x50.8 mm rectangular isolator test section. A blockage cone on a high-speed ball screw actuator downstream of the isolator is pulsed to simulate the throttling behavior in an operational combustor. Wall flush high-speed static pressure transducers detect the streamwise shock train within the isolator. Time delays between the onset of the throttling and the response motion of the shock train are analyzed to assess the controlling scales and parameters. From this analysis, the response time of the isolator shock train system is found to depend primarily on the isolator back pressure time-rate-of-change but quickly saturates to a maximum propagation speed along the isolator. These results inform dual-mode scramjet engine designers of the critical parameters and limits of an unstart resistant engine.

scramjet↗
DOE OSTI2021

Robustness of deep learning algorithms in astronomy -- galaxy morphology studies

Deep learning models are being increasingly adopted in wide array of scientific domains, especially to handle high-dimensionality and volume of the scientific data. However, these models tend to be brittle due to their complexity and overparametrization, especially to the inadvertent adversarial perturbations that can appear due to common image processing such as compression or blurring that are often seen with real scientific data. It is crucial to understand this brittleness and develop models robust to these adversarial perturbations. To this end, we study the effect of observational noise from the exposure time, as well as the worst case scenario of a one-pixel attack as a proxy for compression or telescope errors on performance of ResNet18 trained to distinguish between galaxies of different morphologies in LSST mock data. We also explore how domain adaptation techniques can help improve model robustness in case of this type of naturally occurring attacks and help scientists build more trustworthy and stable models.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

Evaluation of Drilling Performance at The Geysers with Machine Learning Methods Using Geologic Data

A recent well, GDC-36, was drilled in The Geysers Geothermal Field served in a Department of Energy-industry to demonstrate improved drilling performance with polycrystalline diamond compact (PDC) bits. Both PDC and roller cone drill bits were used to drill this well. Key challenges encountered during drilling included lost circulation in the mud-drilled section, and bit damage interfacial severity in the deeper, air-drilled section. The objective of this study is to evaluate the drilling performance in relation to the local geological characteristics using machine learning methods. By applying K-clustering to the sonic log data, we were able to identify areas correlated with measured lost circulation. Also, the boundaries defined by clustering of the mineralogical and lithological data from the mud logs correlate well with interfacial severity during drilling. A random forest model was employed to build correlation between drilling data and rock strength. The confined compressive strength (CCS) of the rock in the training of the machine learning model was inferred from the dipole sonic log. The R-squared of the testing data is 0.78, and the RMSE (Root Mean Squared Error) is 0.06. The trained model was used to forecast rock strength for the section where sonic log data are not available. CCS could also be inferred from mud logs provided the relationship between mineralogy and rock strength is established through core testing data.

15 GEOTHERMAL ENERGY↗
DOE OSTI2022

Nonnegative canonical tensor decomposition with linear constraints: nnCANDELINC

Abstract There is an emerging interest for tensor factorization applications in big‐data analytics and machine learning. To speed up the factorization of extra‐large datasets, organized in multidimensional arrays (also known as tensors), easy to compute compression‐based tensor representations, such as, Tucker and tensor train formats, are used to approximate the initial large‐tensor. Further, tensor factorization is used to extract latent features that can facilitate discoveries of new mechanisms and signatures hidden in the data, where the explainability of the latent features is of principal importance. Nonnegative tensor factorization extracts latent features that are naturally sparse and parts of the data, which makes them easily interpretable. However, to take into account available domain knowledge and subject matter expertise, often additional constraints need to be imposed, which lead us to canonical decomposition with linear constraints (CANDELINC), a canonical polyadic decomposition with rank deficient factors. In CANDELINC, Tucker compression is used as a preprocessing step, which lead to a larger residual error but to more explainable latent features. Here, we propose a nonnegative CANDELINC (nnCANDELINC) accomplished via a specific nonnegative Tucker decomposition; we refer to as minimal or canonical nonnegative Tucker. We derive several results required to understand the specificity of nnCANDELINC, focusing on the difficulties of preserving the nonnegative rank of a tensor to its Tucker core and comparing the real valued to nonnegative case. Finally, we demonstrate nnCANDELINC performance on synthetic and real‐world examples.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1993

Turbulence modeling for high speed compressible flows

The following grant objectives were delineated in the proposal to NASA: to offer course work in computational fluid dynamics (CFD) and related areas to enable mechanical engineering students at North Carolina A&T State University (N.C. A&TSU) to pursue M.S. studies in CFD, and to enable students and faculty to engage in research in high speed compressible flows. Since no CFD-related activity existed at N.C. A&TSU before the start of the NASA grant period, training of students in the CFD area and initiation of research in high speed compressible flows were proposed as the key aspects of the project. To that end, graduate level courses in CFD, boundary layer theory, and fluid dynamics were offered. This effort included initiating a CFD course for graduate students. Also, research work was performed on studying compressibility effects in high speed flows. Specifically, a modified compressible dissipation model, which included a fourth order turbulent Mach number term, was incorporated into the SPARK code and verified for the air-air mixing layer case. The results obtained for this case were compared with a wide variety of experimental data to discern the trends in the mixing layer growth rates with varying convective Mach numbers. Comparison of the predictions of the study with the results of several analytical models was also carried out. The details of the research study are described in the publication entitled 'Compressibility Effects in Modeling Turbulent High Speed Mixing Layers,' which is attached to this report.

Chandra, Suresh↗
NASA NTRS1994

High performance compression of science data

Two papers make up the body of this report. One presents a single-pass adaptive vector quantization algorithm that learns a codebook of variable size and shape entries; the authors present experiments on a set of test images showing that with no training or prior knowledge of the data, for a given fidelity, the compression achieved typically equals or exceeds that of the JPEG standard. The second paper addresses motion compensation, one of the most effective techniques used in the interframe data compression. A parallel block-matching algorithm for estimating interframe displacement of blocks with minimum error is presented. The algorithm is designed for a simple parallel architecture to process video in real time.

Storer, James A.↗
NASA NTRS1994

High Performance Compression of Science Data

Two papers make up the body of this report. One presents a single-pass adaptive vector quantization algorithm that learns a codebook of variable size and shape entries; the authors present experiments on a set of test images showing that with no training or prior knowledge of the data, for a given fidelity, the compression achieved typically equals or exceeds that of the JPEG standard. The second paper addresses motion compensation, one of the most effective techniques used in interframe data compression. A parallel block-matching algorithm for estimating interframe displacement of blocks with minimum error is presented. The algorithm is designed for a simple parallel architecture to process video in real time.

Storer, James A.↗
DOE OSTI2025

Direct interpolative construction of the discrete Fourier transform as a matrix product operator

The quantum Fourier transform (QFT), which can be viewed as a reindexing of the discrete Fourier transform (DFT), has been shown to be compressible as a low-rank matrix product operator (MPO) or quantized tensor train (QTT) operator. However, the original proof of this fact does not furnish a construction of the MPO with a guaranteed error bound. Meanwhile, the existing practical construction of this MPO, based on the compression of a quantum circuit, is not as efficient as possible. We present a simple closed-form construction of the QFT MPO using the interpolative decomposition, with guaranteed near-optimal compression error for a given rank. This construction can speed up the application of the QFT and the DFT, respectively, in quantum circuit simulations and QTT applications. We also connect our interpolative construction to the approximate quantum Fourier transform (AQFT) by demonstrating that the AQFT can be viewed as an MPO constructed using a different interpolation scheme.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Utilizing computer vision and artificial intelligence algorithms to predict and design the mechanical compression response of direct ink write 3D printed foam replacement structures

Additive Manufacturing (AM) of porous polymeric materials, such as foams, recently became a topic of intensive research due their unique combination of low density, impressive mechanical properties, and stress dissipation capabilities. Conventional methods for fabricating foams rely on complex and stochastic processes, making it challenging to achieve precise architectural control of structured porosity. In contrast, AM provides access to a wide range of printable materials, where precise spatial control over structured porosity can be modulated during the fabrication process enabling the production of foam replacement structures (FRS). Current approaches for designing FRS are based on intuitive understanding of their properties or an extensive number of finite element method (FEM) simulations. These approaches, however, are computationally expensive and time consuming. As such, in this work, we present a novel methodology for determining the mechanical compression response of direct ink write (DIW) 3D printed FRS using a simple cross-sectional image. By obtaining measurement data for a relatively small number of samples, an artificial neural network (ANN) was trained, and a computer vision algorithm was used to make inferences about foam compression characteristics from a single cross-sectional image. Finally, a genetic algorithm (GA) was used to solve the inverse design problem, generating the AM printing parameters that an engineer should use to achieve a desired compression response from a DIW printed FRS. The methods developed herein present an avenue for entirely autonomous design and analysis of additively manufactured structures using artificial intelligence.

36 MATERIALS SCIENCE↗
DOE OSTI2025

A Variational Autoencoder Model Toward Molecular Structure Representation Learning of Fuels

Here, in this work, a Variational Autoencoder (VAE)-based data-driven modeling framework is developed with the overarching goal of enabling fuel design. The VAE model is trained on a large dataset with several chemical species to learn a compressed latent space molecular representation. Chemical structure in the form of Simplified Molecular Input Line Entry System (SMILES) string is fed as input, encoded into the VAE latent space, and decoded back to the SMILES string using Long Short-Term Memory (LSTM) networks. Complexities of the VAE training loss function are thoroughly examined by varying the weightage (beta (𝜷) parameter) of the latent space regularization term, thereby assessing the balance between reconstruction accuracy and validity, and focusing on both accurate molecular structure reconstruction and latent space consistency. Two different strategies for 𝜷 variation are evaluated: linear annealing and cyclic annealing. In addition, the impact of total correlation adjustment and hierarchical priors is also studied with regard to the balance between reconstruction fidelity and latent space regularization, and potential issues such as posterior collapse, over-regularization, and poor disentanglement of latent variables. Overall, the best performance of the model is achieved with hierarchical priors and incrementally increasing 𝜷 from 0 to a threshold value of 0.25 over 75 epochs. The generative VAE model can be readily coupled with Quantitative Structure–Property Relationship (QSPR) analysis to develop an integrated end-to-end framework for fuel-property prediction and molecular design of novel promising fuels.

fuel design↗
DOE OSTI2021

Fast convolutional neural networks on FPGAs with hls4ml

We introduce an automated tool for deploying ultra low-latency, low-power deep neural networks with convolutional layers on field-programmable gate arrays (FPGAs). By extending the hls4ml library, we demonstrate an inference latency of 5 µs using convolutional architectures, targeting microsecond latency applications like those at the CERN Large Hadron Collider. Considering benchmark models trained on the Street View House Numbers Dataset, we demonstrate various methods for model compression in order to fit the computational constraints of a typical FPGA device used in trigger and data acquisition systems of particle detectors. In particular, we discuss pruning and quantization-aware training, and demonstrate how resource utilization can be significantly reduced with little to no loss in model accuracy. We show that the FPGA critical resource consumption can be reduced by 97% with zero loss in model accuracy, and by 99% when tolerating a 6% accuracy degradation.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2026

Surrogate Neural Architecture Codesign Package (SNAC-Pack)

Neural architecture search (NAS) is a powerful approach for automating model design, but existing methods often optimize for accuracy alone or rely on proxy metrics such as bit operations (BOPs) that correlate poorly with hardware cost. This gap is particularly large for FPGA deployment, where cost is dominated by a multi-dimensional budget of lookup tables, DSPs, flip-flops, BRAM, and latency. We present the Surrogate Neural Architecture Codesign Package (SNAC-Pack), an open-source AutoML framework for hardware-aware neural architecture codesign and end-to-end FPGA deployment. SNAC-Pack runs a multi-objective global search with Optuna and NSGA-II, loading trials to a shared SQLite store that enables parallel workers across compute nodes. A hardware surrogate model outputs per-trial resource and latency estimates, avoiding the synthesis cost that would otherwise dominate the search loop. A local search stage then applies quantization-aware training (QAT) together with iterative magnitude pruning in a combined compression loop, after which the final model is synthesized to FPGA firmware via the hls4ml Python library. A YAML configuration and an optional agentic frontend let users run the pipeline on new datasets without modifying the framework. We demonstrate SNAC-Pack on jet classification at the Large Hadron Collider and superconducting qubit readout, discovering compact architectures that match or exceed strong baselines on the task metric while reducing FPGA resource utilization and, in the qubit readout case, reducing the design space exploration process from months of manual fine-tuning to hours of automated search.

Weitz, Jason [UC, San Diego]↗
NASA NTRSDate not supplied

Modeling Shed Vorticity from Coaxial Blade Interactions

Coaxial counter-rotating rotors operate in a flowfield different from single rotors. Aerodynamic interactions such as blade crossing and shed vorticity result in potential sources of noise and impulsive blade loads. In previous research, the authors simulated two trains of airfoils traveling in opposite directions for specified speeds, airfoil thickness and vertical separation distances, using the compressible Navier-Stokes solver OVERFLOW. Previously, the effects of circulation, thickness, and compressibility were explored. This work continues the previous research by exploring downwash and shed vorticity effects. These phenomena are explored by simulating two trains of eight airfoils vertically separated traveling in opposite directions. The effects of downwash are simulated by introducing a vertical flow. Vorticity shed from the upper train of airfoils is shown to interact with the lower train, affecting the loading on the lower airfoils. Furthermore, viscid and inviscid calculations are performed to further understand the behavior of shed vorticity.

Natasha L Schatzman↗
DOE OSTI2025

Dynamical Low-Rank Compression of Neural Networks with Robustness under Adversarial Attacks

Deployment of neural networks on resource-constrained devices demands models that are both compact and robust to adversarial inputs. However, compression and adversarial robustness often conflict. In this work, we introduce a dynamical low-rank training scheme enhanced with a novel spectral regularizer that controls the condition number of the low-rank core in each layer. This approach mitigates the sensitivity of compressed models to adversarial perturbations without sacrificing clean accuracy. The method is model- and data-agnostic, computationally efficient, and supports rank adaptivity to automatically compress the network at hand. Extensive experiments across standard architectures, datasets, and adversarial attacks show the regularized networks can achieve over 94 compression while recovering or improving adversarial accuracy relative to uncompressed baselines.

Schotthoefer, Steffen [ORNL] (ORCID:00000002156965↗
DOE OSTI2021

Visualization of shocked material instabilities using a fast-framing camera and XFEL four-pulse train

Many questions regarding dynamic materials could be answered by using time-resolved ultra-fast imaging techniques to characterize the physical and chemical behavior of materials in extreme conditions and their evolution on the nanosecond scale. In this work, we perform multi-frame phase-contrast imaging (PCI) of micro-voids in low density polymers under laser-driven shock compression. At the Matter in Extreme Conditions (MEC) Instrument at the Linac Coherent Light Source (LCLS), we used a train of four x-ray free electron laser (XFEL) pulses to probe the evolution of the samples. To visualize the void and shock wave interaction, here, we deployed the Icarus V2 detector to record up to four XFEL pulses, separated by 1-3 nanoseconds. In this work, we image elastic waves interacting with the micro-voids at a pressure of several GPa. Monitoring how the material’s heterogeneities, like micro-voids, dictate its response to a compressive wave is important for benchmarking the performances of inertial confinement fusion energy materials. For the first time in a single sample, we have combined an ultrafast x-ray framing camera and four XFEL pulse train to create an ultrafast movie of micro-void evolution under laser-driven shock compression. Eventually, we hope this technique will resolve the material density as it evolves dynamically under laser shock compression.

fusion↗
NASA NTRS2006

Comparison of Bystander Cardiopulmonary Resuscitation (BCPR) Performance in the Absence and Presence of Timing Devices for Coordinating Delivery of Ventilatory Breaths and Cardiac Compressions in a Model of Adult Cardiopulmonary Arrest

Astronaut crew medical officers (CMO) aboard the International Space Station (ISS) receive 40 hours of medical training during the 18 months preceding each mission. Part of this training ilncludes twoperson cardiopulmonary resuscitation (CPR) per training guidelines from the American Heart Association (AHA). Recent studies concluded that the use of metronomic tones improves the coordination of CPR by trained clinicians. Similar data for bystander or "trained lay people" (e.g. CMO) performance of CPR (BCPR) have been limited. The purpose of this study was to evailuate whether use of timing devices, such as audible metronomic tones, would improve BCPR perfomance by trained bystanders. Twenty pairs of bystanders trained in two-person BCPR performled BCPR for 4 minutes on a simulated cardiopulmonary arrest patient using three interventions: 1) BCPR with no timing devices, 2) BCPR plus metronomic tones for coordinating compression rate only, 3) BCPR with a timing device and metronome for coordinating ventilation and compression rates, respectively. Bystanders were evaluated on their ability to meet international and AHA CPR guidelines. Bystanders failed to provide the recommended number of breaths and number of compressions in the absence of a timing device and in the presence of audible metronomic tones for only coordinating compression rate. Bystanders using timing devices to coordinate both components of BCPR provided the reco number of breaths and were closer to providing the recommended number of compressions compared with the other interventions. Survey results indicated that bystanders preferred to use a metronome for delivery of compressions during BCPR. BCPR performance is improved by timing devices that coordinate both compressions and breaths.

Hurst, Victor, IV↗
DOE OSTI2025

FedEFsz: Fair Cross-Silo Federated Learning System With Error-Bounded Lossy Compression

Cross-Silo federated learning systems have been identified as an efficient approach to scaling DNN training across geographically-distributed data silos to preserve the privacy of the training data. Communication efficiency and fairness are two major issues that need to be both satisfied when federated learning systems are deployed in practice. Simultaneously guaranteeing both of them, however, is exceptionally difficult because simply combining communication reduction and fairness optimization approaches often causes non-converged training or drastic accuracy degradation. Here, to bridge this gap, we propose FedEFsz. On the one hand, it integrates the state-of-the-art error-bounded lossy compressor SZ3 into cross-silo federated learning systems to significantly reduce communication traffic during the training. On the other hand, it achieves a high fairness (i.e., rather consistent model accuracy and performance across different clients) through a carefully designed heuristic algorithm that can tune the error-bound of SZ3 for different clients during the training. Extensive experimental results based on a GPU cluster with 65 GPU cards show that FedEFsz improves the fairness across different benchmarks by up to 60.88% and meanwhile reduces the communication traffic by up to 315×.

Cross-Silo Federated Learning Systems↗