Search NASA⌕ Search

SEARCH · Search NASA

Results for “labeling”

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 19 records
All sourcesNASA NTRSDOE OSTI
NASA NTRS2019

Image Labeler: Label Earth Science Images for Machine Learning

The application of machine learning for image-based classification of earth science phenomena, such as hurricanes, is relatively new. While extremely useful, the techniques used for image-based phenomena classification require storing and managing an abundant supply of labeled images in order to produce meaningful results. Existing methods for dataset management and labeling include maintaining categorized folders on a local machine, a process that can be cumbersome and not scalable. Image Labeler is a fast and scalable web-based tool that facilitates the rapid development of image-based earth science phenomena datasets, in order to aid deep learning application and automated image classification/detection. Image Labeler is built with modern web technologies to maximize the scalability and availability of the platform. It has a user-friendly interface that allows tagging multiple images relatively quickly. Essentially, Image Labeler improves upon existing techniques by providing researchers with a shareable source of tagged earth science images for all their machine learning needs. Here, we demonstrate Image Labeler’s current image extraction and labeling capabilities including supported data sources, spatiotemporal subsetting capabilities, individual project management and team collaboration for large scale projects.

Acharya, Ashish↗
NASA NTRS2000

Hoechst fluorescence intensity can be used to separate viable bromodeoxyuridine-labeled cells from viable non-bromodeoxyuridine-labeled cells

BACKGROUND: 5-Bromo-2'-deoxyuridine (BrdU) is a powerful compound to study the mitotic activity of a cell. Most techniques that identify BrdU-labeled cells require conditions that kill the cells. However, the fluorescence intensity of the membrane-permeable Hoechst dyes is reduced by the incorporation of BrdU into DNA, allowing the separation of viable BrdU positive (BrdU+) cells from viable BrdU negative (BrdU-) cells. METHODS: Cultures of proliferating cells were supplemented with BrdU for 48 h and other cultures of proliferating cells were maintained without BrdU. Mixtures of viable BrdU+ and viable BrdU- cells from the two proliferating cultures were stained with Hoechst 33342. The viable BrdU+ and BrdU- cells were sorted into different fractions from a mixture of BrdU+ and BrdU- cells based on Hoechst fluorescence intensity and the ability to exclude the vital dye, propidium iodide. Subsequently, samples from the original mixture, the sorted BrdU+ cell population, and the sorted BrdU- cell population were immunostained using an anti-BrdU monoclonal antibody and evaluated using flow cytometry. RESULTS: Two mixtures consisting of approximately 55% and 69% BrdU+ cells were sorted into fractions consisting of greater than 93% BrdU+ cells and 92% BrdU- cells. The separated cell populations were maintained in vitro after sorting to demonstrate their viability. CONCLUSIONS: Hoechst fluorescence intensity in combination with cell sorting is an effective tool to separate viable BrdU+ from viable BrdU- cells for further study. The separated cell populations were maintained in vitro after sorting to demonstrate their viability. Copyright 2000 Wiley-Liss, Inc.

Evaluation Studies↗
NASA NTRSDate not supplied

A Hybrid Approach to Labeling Datasets in Earth Science Publications

NASA Data Centers provide the public with thousands of datasets that result in published papers, reports, and conference proceedings. Collecting accurate metrics on usage of these datasets is key to connecting different areas of knowledge and evaluating the datasets’ impact. While most of the datasets have Digital Object Identifiers (DOIs) assigned, most publications do not cite them hampering the automated search of these publications. Instead, articles mention attributes like organization, instrument, mission, variable, or a publication describing the dataset. Often only domain experts can deduce the dataset that was used in the publication text. The lack of a citation slows the spread of information and reduces the research’s impact. With thousands of papers produced each year, an automated means of labeling datasets is critical. This paper explores a hybrid approach of heuristics and a Natural Language Processing (NLP) Named Entity Recognition (NER) model to find and label the datasets used within Earth Science papers. Heuristics are used to produce the labelled sentences and any potential dataset candidates that can be derived from a sentence. The heuristic labels the sentences with the names of mission, instrument, re-analysis models, and science keywords taken from the Global Change Master Directory (GCMD) ontology. Additionally, it uses those labels to generate the dataset citation candidates. If the mission, instrument, and variable are sufficient to create the citation for the dataset the citation and the label the domain expert reviews the output without going through the NLP model. If the extracted label is not sufficient to label the dataset on its own, the sentence and its associated dataset labels will be inputted into the NER model. The model outputs the labeled sentence and the potential dataset candidates with their associated probabilities. The domain expert then reviews the NER model’s output and the correct labels are determined. The newly labelled papers can then be used as additional training data. This creates an iterative process for the approach to continuously improve. Because all the possible mentions are gathered by the model, the domain expert can quickly and easily label the papers resulting in large time savings.

Jacob Atkins↗
NASA NTRS1989

Specific photoaffinity labeling of two plasma membrane polypeptides with an azido auxin

Plasma membrane vesicles were isolated from zucchini (Cucurbita pepo) hypocotyl tissue by aqueous phase partitioning and assessed for homogeneity by the use of membrane-specific enzyme assays. The highly pure (ca. 95%) plasma membrane vesicles maintained a pH differential across the membrane and accumulated a tritiated azido analogue of 3-indoleacetic acid (IAA), 5-azido-[7-3H]IAA ([3H]N3IAA), in a manner similar to the accumulation of [3H]IAA. The association of the [3H]N3IAA with membrane vesicles was saturable and subject to competition by IAA and auxin analogues. Auxin-binding proteins were photoaffinity labeled by addition of [3H]N3IAA to plasma membrane vesicles prior to exposure to UV light (15 sec; 300 nm) and detected by subsequent NaDodSO4/PAGE and fluorography. When the reaction temperature was lowered to -196 degrees C, high-specific-activity labeling of a 40-kDa and a 42-kDa polypeptide was observed. Triton X-100 (0.1%) increased the specific activity of labeling and reduced the background, which suggests that the labeled polypeptides are intrinsic membrane proteins. The labeled polypeptides are of low abundance, as expected for auxin receptors. Further, the addition of IAA and auxin analogues to the photoaffinity reaction mixture resulted in reduced labeling that was qualitatively similar to their effects on the accumulation of radiolabeled IAA in membrane vesicles. Collectively, these results suggest that the radiolabeled polypeptides are auxin receptors. The covalent nature of the label should facilitate purification and further characterization of the receptors.

NASA Discipline Number 29-20↗
NASA NTRS1999

Textural-Contextual Labeling and Metadata Generation for Remote Sensing Applications

Despite the extensive research and the advent of several new information technologies in the last three decades, machine labeling of ground categories using remotely sensed data has not become a routine process. Considerable amount of human intervention is needed to achieve a level of acceptable labeling accuracy. A number of fundamental reasons may explain why machine labeling has not become automatic. In addition, there may be shortcomings in the methodology for labeling ground categories. The spatial information of a pixel, whether textural or contextual, relates a pixel to its surroundings. This information should be utilized to improve the performance of machine labeling of ground categories. Landsat-4 Thematic Mapper (TM) data taken in July 1982 over an area in the vicinity of Washington, D.C. are used in this study. On-line texture extraction by neural networks may not be the most efficient way to incorporate textural information into the labeling process. Texture features are pre-computed from cooccurrence matrices and then combined with a pixel's spectral and contextual information as the input to a neural network. The improvement in labeling accuracy with spatial information included is significant. The prospect of automatic generation of metadata consisting of ground categories, textural and contextual information is discussed.

Kiang, Richard K.↗
NASA NTRS2016

Implementing Connected Component Labeling as a User Defined Operator for SciDB

We have implemented a flexible User Defined Operator (UDO) for labeling connected components of a binary mask expressed as an array in SciDB, a parallel distributed database management system based on the array data model. This UDO is able to process very large multidimensional arrays by exploiting SciDB's memory management mechanism that efficiently manipulates arrays whose memory requirements far exceed available physical memory. The UDO takes as primary inputs a binary mask array and a binary stencil array that specifies the connectivity of a given cell to its neighbors. The UDO returns an array of the same shape as the input mask array with each foreground cell containing the label of the component it belongs to. By default, dimensions are treated as non-periodic, but the UDO also accepts optional input parameters to specify periodicity in any of the array dimensions. The UDO requires four stages to completely label connected components. In the first stage, labels are computed for each subarray or chunk of the mask array in parallel across SciDB instances using the weighted quick union (WQU) with half-path compression algorithm. In the second stage, labels around chunk boundaries from the first stage are stored in a temporary SciDB array that is then replicated across all SciDB instances. Equivalences are resolved by again applying the WQU algorithm to these boundary labels. In the third stage, relabeling is done for each chunk using the resolved equivalences. In the fourth stage, the resolved labels, which so far are "flattened" coordinates of the original binary mask array, are renamed with sequential integers for legibility. The UDO is demonstrated on a 3-D mask of O(1011) elements, with O(108) foreground cells and O(106) connected components. The operator completes in 19 minutes using 84 SciDB instances.

UDO↗
NASA NTRS1976

Studies related to the development of the Viking 1975 labeled release experiment

The labeled release life detection experiment on the Viking 1975 Mars mission is based on the concept that microorganisms will metabolize radioactive organic substrates in a nutrient medium and release radioactive carbon dioxide. Several experiments, using laboratory equipment, were carried out to evaluate various aspects of the concept. Results indicate: (1) label is released by sterilization-treated soil, (2) substantial quantities of label are retained in solution under basic conditions, (3) the substrate used, as well as position of label in the molecule, affect release of label, (4) label release is depressed by radiolytic decomposition of substrates, and (5) About 100,000 organisms are required to produce a detectable response. These results, suggest additional areas for testing, add to the data base for interpretation of flight results, and have significance for broader application of this technique for assessing microbial activity.

Devincenzi, D. L.↗
NASA NTRS2008

Gold Nanoparticle Labels Amplify Ellipsometric Signals

The ellipsometric method reported in the immediately preceding article was developed in conjunction with a method of using gold nanoparticles as labels on biomolecules that one seeks to detect. The purpose of the labeling is to exploit the optical properties of the gold nanoparticles in order to amplify the measurable ellipsometric effects and thereby to enable ultrasensitive detection of the labeled biomolecules without need to develop more-complex ellipsometric instrumentation. The colorimetric, polarization, light-scattering, and other optical properties of nanoparticles depend on their sizes and shapes. In the present method, these size-and-shape-dependent properties are used to magnify the polarization of scattered light and the diattenuation and retardance of signals derived from ellipsometry. The size-and-shape-dependent optical properties of the nanoparticles make it possible to interrogate the nanoparticles by use of light of various wavelengths, as appropriate, to optimally detect particles of a specific type at high sensitivity. Hence, by incorporating gold nanoparticles bound to biomolecules as primary or secondary labels, the performance of ellipsometry as a means of detecting the biomolecules can be improved. The use of gold nanoparticles as labels in ellipsometry has been found to afford sensitivity that equals or exceeds the sensitivity achieved by use of fluorescence-based methods. Potential applications for ellipsometric detection of gold nanoparticle-labeled biomolecules include monitoring molecules of interest in biological samples, in-vitro diagnostics, process monitoring, general environmental monitoring, and detection of biohazards.

Venkatasubbarao, Srivatsa↗
NASA NTRS2005

Digital Equivalent Data System for XRF Labeling of Objects

A digital equivalent data system (DEDS) is a system for identifying objects by means of the x-ray fluorescence (XRF) spectra of labeling elements that are encased in or deposited on the objects. As such, a DEDS is a revolutionary new major subsystem of an XRF system. A DEDS embodies the means for converting the spectral data output of an XRF scanner to an ASCII alphanumeric or barcode label that can be used to identify (or verify the assumed or apparent identity of) an XRF-scanned object. A typical XRF spectrum of interest contains peaks at photon energies associated with specific elements on the Periodic Table (see figure). The height of each spectral peak above the local background spectral intensity is proportional to the relative abundance of the corresponding element. Alphanumeric values are assigned to the relative abundances of the elements. Hence, if an object contained labeling elements in suitably chosen proportions, an alphanumeric representation of the object could be extracted from its XRF spectrum. The mixture of labeling elements and for reading the XRF spectrum would be compatible with one of the labeling conventions now used for bar codes and binary matrix patterns (essentially, two-dimensional bar codes that resemble checkerboards). A further benefit of such compatibility is that it would enable the conversion of the XRF spectral output to a bar or matrix-coded label, if needed. In short, a process previously used only for material composition analysis has been reapplied to the world of identification. This new level of verification is now being used for "authentication."

Schramm, Harry F.↗
NASA NTRS2011

Using Ensemble Decisions and Active Selection to Improve Low-Cost Labeling for Multi-View Data

This paper seeks to improve low-cost labeling in terms of training set reliability (the fraction of correctly labeled training items) and test set performance for multi-view learning methods. Co-training is a popular multiview learning method that combines high-confidence example selection with low-cost (self) labeling. However, co-training with certain base learning algorithms significantly reduces training set reliability, causing an associated drop in prediction accuracy. We propose the use of ensemble labeling to improve reliability in such cases. We also discuss and show promising results on combining low-cost ensemble labeling with active (low-confidence) example selection. We unify these example selection and labeling strategies under collaborative learning, a family of techniques for multi-view learning that we are developing for distributed, sensor-network environments.

machine learning↗
NASA NTRS1979

Learning with imperfectly labeled patterns

The problem of learning in pattern recognition using imperfectly labeled patterns is considered. The performance of the Bayes and nearest neighbor classifiers with imperfect labels is discussed using a probabilistic model for the mislabeling of the training patterns. Schemes for training the classifier using both parametric and non parametric techniques are presented. Methods for the correction of imperfect labels were developed. To gain an understanding of the learning process, expressions are derived for success probability as a function of training time for a one dimensional increment error correction classifier with imperfect labels. Feature selection with imperfectly labeled patterns is described.

Chittineni, C. B.↗
NASA NTRS1980

Label Identification from Statistical Tabulation (LIST) temporal extendability study

The author has identified the following significant results. The most apparent contributors to the problem of poor temporal extension of LIST are the drastic changes in the brightness keys and an inadequate set of AI responses in Phase 3. The brightness trajectories change drastically from Phase 3 to the transition year (TY). Removing brightness channels from the discriminant does not completely correct the lack of extendability. Removing brightness increases the accuracy of the extension from Phase 3 to TY from 57.7 percent to 64.18 percent. The removal of the Al keys increases accuracy to 65.76 percent. Although the latter increase appears insignificant when compared to the first, the removal of only the Al keys increased accuracy to 63.58 percent. Proper weighting of the responses explains 73.8 percent of the ground truth labels but only 56.7 percent of the Al labels. By contrast, the TY responses which were weighted to explain the TY ground truth labels fared equally well, explaining 73.6 percent of those labels and 87.1 percent of the Al labels.

Dennis, T. B.↗
NASA NTRS1980

Some approaches to optimal cluster labeling of aerospace imagery

Some approaches are presented to the problem of labeling clusters using information from a given set of labeled and unlabeled aerospace imagery patterns. The assignment of class labels to the clusters is formulated as the determination of the best assignment over all possible ones with respect to some criterion. Cluster labeling is also viewed as the probability of correct labeling with a maximization of likelihood function. Results of the application of these techniques in the processing of remotely sensed multispectral scanner imagery data are presented.

Chittineni, C. B.↗
NASA NTRS1981

On the accuracy of pixel relaxation labeling

An analysis of pixel labeling by probabilistic relaxation techniques is presented to demonstrate that these labeling procedures degenerate to weighted averages in the vicinity of fixed points. A consequence of this is that undesired label conversions can occur, leading to a deterioration of labeling accuracy at a stage after an improvement has already been achieved. Means for overcoming the accuracy deterioration are suggested and used as the basis for a possible design strategy for using probabilistic relaxation procedures. The results obtained are illustrated using simple data sets in which labeling on individual pixels can be examined and also using Landsat imagery to show application to data typical of that encountered in remote sensing applications.

Richards, J. A.↗
NASA NTRS1980

Maximum likelihood estimation of label imperfection probabilities and its use in the identification of mislabeled patterns

Estimating label imperfections and the use of estimations in the identification of mislabeled patterns are discussed. Expressions are presented for the asymptotic variances of the probability of correct classification and proportion, and for the maximum likelihood estimates of classification errors and a priori probabilities. Models are developed for imperfections in the labels and classification errors, and expressions are derived for the probability of imperfect label identification schemes resulting in wrong decisions. The expressions are used in computing thresholds and the techniques are given practical applications. The imperfect label identification scheme in the multiclass case is found to amount to establishing a region around each decision surface, and decisions of the label correction scheme are found in close agreement with the analyst-interpreter interpretations of the imagery films. As an example, the application of the maximum likelihood estimation to the processing of Landsat MSS data is discussed.

Chittineni, C. B.↗