Search NASA⌕ Search

SEARCH · Search NASA

Results for “vector quantization”

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
DOE OSTI2021

Attention Network Forecasts Time-to-Failure in Laboratory Shear Experiments

Rocks under stress deform by creep mechanisms that include formation and slip on small-scale internal cracks. Intragranular cracks and slip along grain contacts release energy as elastic waves termed acoustic emissions (AE). AEs are thought to contain predictive information that can be used for fault failure forecasting. Here, we present a method using unsupervised classification and an attention network to forecast labquakes using AE waveform features. Our data were generated in a laboratory setting using a biaxial shearing device with granular fault gouge intended to mimic the conditions of tectonic faults. Here, we analyzed the temporal evolution of AEs generated throughout several hundred laboratory earthquake cycles. We used a Conscience Self-Organizing Map (CSOM) to perform topologically ordered vector quantization based on waveform properties. The resulting map was used to interactively cluster AEs. We examined the clusters over time to identify those with predictive ability. Finally, we used a variety of LSTM and attention-based networks to test the predictive power of the AE clusters. By tracking cumulative waveform features over the seismic cycle, the network is able to forecast the time-to-failure (TTF) of lab earthquakes. Our results show that analyzing the data to isolate predictive signals and using a more sophisticated network architecture are key to robustly forecasting labquakes. In the future, this method could be applied on tectonic faults to monitor earthquakes and augment early warning systems.

58 GEOSCIENCES↗
DOE OSTI2023

Comparing storm resolving models and climates via unsupervised machine learning

Global storm-resolving models (GSRMs) have gained widespread interest because of the unprecedented detail with which they resolve the global climate. However, it remains difficult to quantify objective differences in how GSRMs resolve complex atmospheric formations. This lack of comprehensive tools for comparing model similarities is a problem in many disparate fields that involve simulation tools for complex data. To address this challenge we develop methods to estimate distributional distances based on both nonlinear dimensionality reduction and vector quantization. Our approach automatically learns physically meaningful notions of similarity from low-dimensional latent data representations that the different models produce. This enables an intercomparison of nine GSRMs based on their high-dimensional simulation data (2D vertical velocity snapshots) and reveals that only six are similar in their representation of atmospheric dynamics. Furthermore, we uncover signatures of the convective response to global warming in a fully unsupervised way. Our study provides a path toward evaluating future high-resolution simulation data more objectively.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

Masked Particle Modeling on Sets: Towards Self-Supervised High Energy Physics Foundation Models

Abstract We propose masked particle modeling (MPM) as a self-supervised method for learning generic, transferable, and reusable representations on unordered sets of inputs for use in high energy physics (HEP) scientific data. This work provides a novel scheme to perform masked modeling based pre-training to learn permutation invariant functions on sets. More generally, this work provides a step towards building large foundation models for HEP that can be generically pre-trained with self-supervised learning and later fine-tuned for a variety of down-stream tasks. In MPM, particles in a set are masked and the training objective is to recover their identity, as defined by a discretized token representation of a pre-trained vector quantized variational autoencoder. We study the efficacy of the method in samples of high energy jets at collider physics experiments, including studies on the impact of discretization, permutation invariance, and ordering. We also study the fine-tuning capability of the model, showing that it can be adapted to tasks such as supervised and weakly supervised jet classification, and that the model can transfer efficiently with small fine-tuning data sets to new classes and new data domains.

Heinrich, Lukas (ORCID:0000000240487584)↗
DOE OSTI2021

Beyond the hubble sequence – exploring galaxy morphology with unsupervised machine learning

We explore unsupervised machine learning for galaxy morphology analyses using a combination of feature extraction with a vector-quantized variational autoencoder (VQ-VAE) and hierarchical clustering (HC). We propose a new methodology that includes: (1) consideration of the clustering performance simultaneously when learning features from images; (2) allowing for various distance thresholds within the HC algorithm; (3) using the galaxy orientation to determine the number of clusters. This set-up provides 27 clusters created with this unsupervised learning that we show are well separated based on galaxy shape and structure (e.g. Sérsic index, concentration, asymmetry, Gini coefficient). These resulting clusters also correlate well with physical properties such as the colour–magnitude diagram, and span the range of scaling relations such as mass versus size amongst the different machine-defined clusters. When we merge these multiple clusters into two large preliminary clusters to provide a binary classification, an accuracy of $\sim 87{{\ \rm per\ cent}}$ is reached using an imbalanced data set, matching real galaxy distributions, which includes 22.7 per cent early-type galaxies and 77.3 per cent late-type galaxies. Comparing the given clusters with classic Hubble types (ellipticals, lenticulars, early spirals, late spirals, and irregulars), we show that there is an intrinsic vagueness in visual classification systems, in particular galaxies with transitional features such as lenticulars and early spirals. Based on this, the main result in this work is not how well our unsupervised method matches visual classifications and physical properties, but that the method provides an independent classification that may be more physically meaningful than any visually based ones.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2025

Knowledge Oriented Graph Unified Transformer (KOGUT) v0.1

KOGUT — Knowledge Oriented Graph Unified Transformer KOGUT implements the Relational Graph Transformer (RelGT) architecture for knowledge graph link prediction in biological domains, with a primary focus on microbial growth media prediction. While the original RelGT (arXiv:2505.10960) targets relational tables, time series, and multi-table databases, KOGUT adapts this architecture for heterogeneous biological knowledge graphs, providing first-in-class AI predictive models for microbial cultivation. Key Adaptations Beyond Original RelGT: - Knowledge Graph Focus: Applied to biological KGs with semantic node types (taxa, chemicals, media, phenotypes, environments) versus generic relational database tables, trained on the KG-Microbe knowledge graph (1.3M entities, 2.9M edges, 24 relation types). - Multimodal Node Encoding: Integrates node labels, categories, descriptions, and synonyms from KG metadata through learned embedding layers—adapting relational column features to graph node attributes with textual semantics. - Extended K-Hop Subgraph Strategy: Optimized neighborhood sampling (3-hop default, configurable up to 200 nodes) tuned for sparse biological networks, building on the original local-global attention framework with biological relation preservation. - Biolink Predicate Preservation: Type-specific transformations for 24 biological edge semantics (occurs_in, consumes, produces, has_phenotype, subclass_of) beyond standard relational foreign keys, enabling multi-relation link prediction. - Inductive Learning Support: Enables zero-shot predictions for novel taxa through feature-based embeddings (temperature, oxygen requirements, gram stain, cell shape), extending the original transductive relational benchmark scope to uncultured microorganisms. CheapSOTA Performance Optimizations (This Distribution): - VQ-EMA Centroid Attention: Vector quantization with exponential moving average for improved global context modeling (+5-10% MRR improvement). - HDF5 Precomputed Data Loading: One-time preprocessing of k-hop subgraphs to eliminate redundant graph traversals (2-5× training speedup). - Distributed Data Parallel Training: Multi-GPU support for scaling to larger knowledge graphs (tested on 4× NVIDIA A100 GPUs at NERSC Perlmutter). - Mixed Precision Training: Automatic mixed precision (AMP) for memory efficiency and faster training. Advantages Over Standard Knowledge Graph Embedding Models: Combines RelGT's proven multi-element tokenization (features, type, hop, structure) with graph-native biological representations, enabling interpretable link prediction across heterogeneous entities that standard embedding models (TransE, RotatE, ComplEx) and table-based transformers cannot directly model. Achieves near-perfect performance on microbial growth media prediction (MRR: 0.9966, Precision@1: 0.9932, Hit@10: 1.0000) while maintaining explainability through attention-based reasoning over biological pathways. Training Data: - KG-Microbe merged knowledge graph: 1,379,337 nodes, 2,960,472 edges - 24 biological relation types including taxonomic hierarchies, metabolic interactions, phenotype associations, and environmental relationships - Primary prediction task: Growth media suitability for microbial taxa (biolink:occurs_in, 50K edges) - Multi-relation capability: Predicts links for any of the 24 relation types, including chemical consumption/production, phenotype associations, and taxonomic classification Citation: Original RelGT Architecture: Dwivedi et al., "Relational Graph Transformer", arXiv:2505.10960, 2025 KOGUT Implementation: Knowledge Oriented Graph Unified Transformer for Microbial Growth Media Prediction Developed at Lawrence Berkeley National Laboratory (LBNL) Trained on NERSC Perlmutter supercomputer

Joachimiak, Marcin [Lawrence Berkeley National Lab↗
DOE OSTI2025

Electron-Proton Scattering Event Generation using Structured Tokenization

Recent work such as Omnijet-$\alpha$ has demonstrated that effective tokenization combined with transformer-based architectures can produce effective foundation models for jet physics. While tokenization may help models capture generalizable event characteristics, it also introduces discretization errors that may compromise the precision required for downstream physics analyses. As the number and complexity of the particle features grow, these errors are likely to grow proportionally. In this study, we investigate new tokenization strategies to improve the application of generative transformer models to \textsc{Pythia8} simulations of electron-proton scattering at the Electron-Ion Collider. Specifically, we propose a feature-based structured tokenization approach that utilizes multiple tokens per particle, improving expressivity, while reducing the total number of unique tokens needed. We evaluate this method against grid-based binning, K-means clustering, and vector-quantized variational auto-encoders on the event simulations. Our results show that feature-based structured tokenization reduces discretization error, leading to more accurate generative modeling of particle-level events.

Goldenberg, Steven [Thomas Jefferson National Acce↗
DOE OSTI2021

Overview of Algorithms for Using Particle Morphology in Pre-Detonation Nuclear Forensics

A major goal in pre-detonation nuclear forensics is to infer the processing conditions and/or facility type that produced radiological material. This review paper focuses on analyses of particle size, shape, texture (“morphology”) signatures that could provide information on the provenance of interdicted materials. For example, uranium ore concentrates (UOC or yellowcake) include ammonium diuranate (ADU), ammonium uranyl carbonate (AUC), sodium diuranate (SDU), magnesium diuranate (MDU), and others, each prepared using different salts to precipitate U from solution. Once precipitated, UOCs are often dried and calcined to remove adsorbed water. The products can be allowed to react further, forming uranium oxides UO3, U3O8, or UO2 powders, whose surface morphology can be indicative of precipitation and/or calcination conditions used in their production. This review paper describes statistical issues and approaches in using quantitative analyses of measurements such as particle size and shape to infer production conditions. Statistical topics include multivariate t tests (Hotelling’s T 2 ), design of experiments, and several machine learning (ML) options including decision trees, learning vector quantization neural networks, mixture discriminant analysis, and approximate Bayesian computation (ABC). ABC is emphasized as an attractive option to include the effects of model uncertainty in the selected and fitted forward model used for inferring processing conditions.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗
NASA NTRS1987

Parallel image compression

A parallel compression algorithm for the 16,384 processor MPP machine was developed. The serial version of the algorithm can be viewed as a combination of on-line dynamic lossless test compression techniques (which employ simple learning strategies) and vector quantization. These concepts are described. How these concepts are combined to form a new strategy for performing dynamic on-line lossy compression is discussed. Finally, the implementation of this algorithm in a massively parallel fashion on the MPP is discussed.

Reif, John H.↗
NASA NTRS1988

The design and performance of a real-time self excited vocoder

This paper is concerned with a generic class of predictive speech coders that includes the newly proposed Self Excited Vocoder (SEV) and the well known Code-Excited Linear Predictive Coder (CELPC). All members of this class form an excitation sequence for a linear predictive model filter using the same general model for the excitation signal. The general excitation model is based on a block coding technique where each sequence is drawn from an ensemble of sequences. This paper reports on two developments related to this general model. The first development is a new type of excitation ensemble that can in general be populated by many different types of sequences. The second development is a means of populating this new type of ensemble based on a vector quantizer design procedure using a new distortion measure.

Rose, Richard C.↗
NASA NTRS1988

Spatial compression of Seasat SAR imagery

The results of a study of techniques for spatial compression of synthetic-aperture-radar (SAR) imagery are summarized. Emphasis is on image-data volume reduction for archive and online storage applications while preserving the image resolution and radiometric fidelity. A quantitative analysis of various techniques, including vector quantization (VQ) and adaptive discrete cosine transform (ADCT), is presented. Various factors such as compression ratio, algorithm complexity, and image quality are considered in determining the optimal algorithm. The compression system requirements are established for electronic access of an online archive system based on the results of a survey of the science community. The various algorithms are presented and their results evaluated considering the effects of speckle noise and the wide dynamic range inherent in SAR imagery. The conclusion is that although the ADCT produces the best signal-to-distortion-noise ratio for a given compression ratio, the two-level tree-searched VQ technique is preferred due to its simplicity of decoding and near-optimal performance.

Chang, C. Y.↗
NASA NTRS1989

Data compression experiments with LANDSAT thematic mapper and Nimbus-7 coastal zone color scanner data

A case study is presented where an image segmentation based compression technique is applied to LANDSAT Thematic Mapper (TM) and Nimbus-7 Coastal Zone Color Scanner (CZCS) data. The compression technique, called Spatially Constrained Clustering (SCC), can be regarded as an adaptive vector quantization approach. The SCC can be applied to either single or multiple spectral bands of image data. The segmented image resulting from SCC is encoded in small rectangular blocks, with the codebook varying from block to block. Lossless compression potential (LDP) of sample TM and CZCS images are evaluated. For the TM test image, the LCP is 2.79. For the CZCS test image the LCP is 1.89, even though when only a cloud-free section of the image is considered the LCP increases to 3.48. Examples of compressed images are shown at several compression ratios ranging from 4 to 15. In the case of TM data, the compressed data are classified using the Bayes' classifier. The results show an improvement in the similarity between the classification results and ground truth when compressed data are used, thus showing that compression is, in fact, a useful first step in the analysis.

Tilton, James C.↗
NASA NTRS1987

SAR image data compression for an on-line archive system

This paper summarizes the investigation of SAR image data compression for an on-line archive data distribution system. This system is planned for the ground processing system of Alaska SAR Facility (ASF) and Shuttle Imaging Radar (SIR-C). The objective of the SAR image data compression is to enable the data archive system to provide the remote users a large data base with good image quality, short response time, low transfer cost, and minimal decoding complexity. The requirements and limitations of the on-line archive data distribution system are presented. The effects of SAR image data characteristics on data compression are addressed. The users' survey results suggest that compression ratios between 10:1 and 20:1 appear suitable. Based on the algorithm evaluation results, the two-level tree-searched vector quantization technique has been recommended as the SAR image data compression algorithm for the on-line archive data distribution system.

Chang, C. Y.↗
NASA NTRS1988

Compression of high spectral resolution imagery

The Earth Orbiting Satellite (EOS), scheduled for launch in the mid-1990s, will include a next-generation multispectral imaging system (HIRIS) having unprecedented spatial and spectral resolution. Its high resolution, however, comes at the cost of a raw data rate which exceeds the communication channel capacity assigned to the entire EOS mission. This paper explores noisy compression algorithms which may compress multispectral data by up to 30:1 or more. Algorithm performance is measured using both traditional and mission-oriented criteria (e.g., feature classification consistency). It is shown that vector quantization, merged with suitable preprocessing techniques, is the most viable candidate.

Baker, Richard L.↗
NASA NTRS1989

Advances in image compression and automatic target recognition; Proceedings of the Meeting, Orlando, FL, Mar. 30, 31, 1989

Various papers on image compression and automatic target recognition are presented. Individual topics addressed include: target cluster detection in cluttered SAR imagery, model-based target recognition using laser radar imagery, Smart Sensor front-end processor for feature extraction of images, object attitude estimation and tracking from a single video sensor, symmetry detection in human vision, analysis of high resolution aerial images for object detection, obscured object recognition for an ATR application, neural networks for adaptive shape tracking, statistical mechanics and pattern recognition, detection of cylinders in aerial range images, moving object tracking using local windows, new transform method for image data compression, quad-tree product vector quantization of images, predictive trellis encoding of imagery, reduced generalized chain code for contour description, compact architecture for a real-time vision system, use of human visibility functions in segmentation coding, color texture analysis and synthesis using Gibbs random fields.

Tescher, Andrew G.↗
NASA NTRS1990

Distributed memory approaches for robotic neural controllers

The suitability is explored of two varieties of distributed memory neutral networks as trainable controllers for a simulated robotics task. The task requires that two cameras observe an arbitrary target point in space. Coordinates of the target on the camera image planes are passed to a neural controller which must learn to solve the inverse kinematics of a manipulator with one revolute and two prismatic joints. Two new network designs are evaluated. The first, radial basis sparse distributed memory (RBSDM), approximates functional mappings as sums of multivariate gaussians centered around previously learned patterns. The second network types involved variations of Adaptive Vector Quantizers or Self Organizing Maps. In these networks, random N dimensional points are given local connectivities. They are then exposed to training patterns and readjust their locations based on a nearest neighbor rule. Both approaches are tested based on their ability to interpolate manipulator joint coordinates for simulated arm movement while simultaneously performing stereo fusion of the camera data. Comparisons are made with classical k-nearest neighbor pattern recognition techniques.

Jorgensen, Charles C.↗
NASA NTRS1989

Data compression for Eos on-board SAR processor

A lightweight, low-power, real-time data compressor design for the Earth Observing System (Eos) onboard synthetic aperture radar (SAR) processor is presented. The implementation is based on VLSI design of a pipelined binary tree-searched vector quantizer (VQ), utilizing space-qualifiable 1.25-micron CMOS technology. The implementation exploits VLSI system design principles such as the modularity, regular data flow, simple interconnection, localized communication, simple global control, and parallel/pipelined processing. The overall system requires 30 chips with only one VLSI processing element design. The total weight is about 1.2 lbs, with an estimated power dissipation of approximately 4 watts operating at the maximum input data rate. The projected throughput rate exceeds 5 MHz.

Chang, C. Y.↗
NASA NTRS1991

SAR data compression: Application, requirements, and designs

The feasibility of reducing data volume and data rate is evaluated for the Earth Observing System (EOS) Synthetic Aperture Radar (SAR). All elements of data stream from the sensor downlink data stream to electronic delivery of browse data products are explored. The factors influencing design of a data compression system are analyzed, including the signal data characteristics, the image quality requirements, and the throughput requirements. The conclusion is that little or no reduction can be achieved in the raw signal data using traditional data compression techniques (e.g., vector quantization, adaptive discrete cosine transform) due to the induced phase errors in the output image. However, after image formation, a number of techniques are effective for data compression.

Curlander, John C.↗
NASA NTRS1991

A progressive data compression scheme based upon adaptive transform coding: Mixture block coding of natural images

A method for efficiently coding natural images using a vector-quantized variable-blocksized transform source coder is presented. The method, mixture block coding (MBC), incorporates variable-rate coding by using a mixture of discrete cosine transform (DCT) source coders. Which coders are selected to code any given image region is made through a threshold driven distortion criterion. In this paper, MBC is used in two different applications. The base method is concerned with single-pass low-rate image data compression. The second is a natural extension of the base method which allows for low-rate progressive transmission (PT). Since the base method adapts easily to progressive coding, it offers the aesthetic advantage of progressive coding without incorporating extensive channel overhead. Image compression rates of approximately 0.5 bit/pel are demonstrated for both monochrome and color images.

Rost, Martin C.↗