Search NASA⌕ Search

SEARCH · Search NASA

Results for “LABELLING”

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 37 records · Page 2
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Facile Labeling of Sieve Element Phloem-Protein Bodies Using the Reciprocal Oligosaccharide Probe OGA488

Sieve elements of many angiosperms contain structural phloem proteins (P-proteins) that can interact to create large P-protein bodies. P-protein bodies can occlude sieve plates upon injury but the range of functional and physiological roles of P-proteins remains uncertain, in part because of challenges in labeling and visualization methods. Here, we show that a reciprocal oligosaccharide probe, OGA 488 , can be used in rapid and sensitive labeling of P-protein bodies in Arabidopsis, poplar, snap bean and cucumber in histological sections. OGA 488 labeling of knockouts of the two Arabidopsis P-protein-encoding genes, AtSEOR1 and AtSEOR2 , indicated that labeling is specific to AtSEOR2. That protein bodies were labeled and visible in Atseor1 knockouts indicates that heterodimerization of AtSEOR1 and AtSEOR2 may not be necessary for P-protein body formation. Double labeling with a previously characterized stain for P-proteins, sulphorhodamine 101, confirmed P-protein labeling and also higher specificity of OGA 488 for P-proteins. OGA 488 is thus robust and easily used to label P-proteins in histological sections of multiple angiosperm species.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2017

Label Transfer from APOGEE to LAMOST: Precise Stellar Parameters for 450,000 LAMOST Giants

In this era of large-scale spectroscopic stellar surveys, measurements of stellar attributes (“labels,” i.e., parameters and abundances) must be made precise and consistent across surveys. Here, we demonstrate that this can be achieved by a data-driven approach to spectral modeling. With The Cannon, we transfer information from the APOGEE survey to determine precise T{sub eff}, log g, [Fe/H], and [α/M] from the spectra of 450,000 LAMOST giants. The Cannon fits a predictive model for LAMOST spectra using 9952 stars observed in common between the two surveys, taking five labels from APOGEE DR12 as ground truth T{sub eff}, log g, [Fe/H], [α/M], and K-band extinction A{sub k}. The model is then used to infer T{sub eff}, log g, [Fe/H], and [α/M] for 454,180 giants, 20% of the LAMOST DR2 stellar sample. These are the first [α/M] values for the full set of LAMOST giants, and the largest catalog of [α/M] for giant stars to date. Furthermore, these labels are by construction on the APOGEE label scale; for spectra with S/N > 50, cross-validation of the model yields typical uncertainties of 70 K in T{sub eff}, 0.1 in log g, 0.1 in [Fe/H], and 0.04 in [α/M], values comparable to the broadly stated, conservative APOGEE DR12 uncertainties. Thus, by using “label transfer” to tie low-resolution (LAMOST R ≈ 1800) spectra to the label scale of a much higher-resolution (APOGEE R ≈ 22,500) survey, we substantially reduce the inconsistencies between labels measured by the individual survey pipelines. This demonstrates that label transfer with The Cannon can successfully bring different surveys onto the same physical scale.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Label Assist: Personalized Travel Models for Longitudinal Data Collection

Understanding travel behavior is crucial to transportation decarbonization. OpenPATH is an open-source mobility platform which collects and analyzes human travel behavior at the individual level. The mobile application passively senses trips and prompts users to label them. However, users find the labeling process burdensome; less than half the trips are typically labeled, making much of the data unusable in aggregate analyses of mobility patterns. Prior work has addressed the response fatigue challenge through automated mode inference using sensor data, but sensors cannot capture all aspects of travel behavior. We explore an alternative approach in which we leverage prior user input to predict travel choices in novel trips. We first explore trip clustering methods and develop a novel two-step pipeline using DBSCAN and SVMs to extract realistic geospatial clusters. We then propose two strategies to predict trip labels: (i) clustering trips and extrapolating labels for similar trips, and (ii) random forest classification. The random forest approach is able to achieve - $70-80% accuracy (purpose: 72%, mode: 79%, replaced mode: 81%). These novel approaches to trip classification allow us to increase the rate of user labeling by suggesting predicted labels to be verified by the user. Unlabeled trips can also contribute to aggregate analyses, using label predictions and their associated confidences as a substitute. While there exist other travel survey apps with the ability to infer travel choices, to our knowledge, this is the first paper to describe such a supervised system and rigorously evaluate it.

ADVANCED PROPULSION SYSTEMS,ENERGY PLANNING, POLIC↗
DOE OSTI2023

Syntheses and characterization of isotopically labeled 1,3,5-triamino-2,4,6-trinitrobenzene (TATB)

Synthesis and characterization of chemical analogues of TATB, where specific atoms in the structure have been isotopically substituted, are reported. 15 N, 2 H, and 18 O have replaced the naturally occurring isotope distributions in the amino and/or the nitro attendant sites and 13 C has replaced the carbon in the ring structure. A modified wet-amination method was used to produce the analogues, and the isotopic replacements were performed by selective choice of labeled precursors. Four 15 N-labeled compounds (N replaced in the amino and nitro positions), two deuterium-labeled compounds (hydrogens replaced on the amino groups), and one 13 C-labeled compound (C in the ring substituted) were synthesized of high isotopic and chemical purity. One partially labeled 18 O-labeled compound (O in the nitro position) was a result of incomplete labeling due to exchange reactions during synthesis. The compounds were characterized by various spectroscopic methods – mass spectrometry (MS), solid-state nuclear magnetic resonance (SS-NMR), infrared (FTIR), powder x-ray diffraction (PXRD), and differential scanning calorimetry (DSC), depending upon the substitution. In conclusion, these compounds have been critical to the efforts in understanding the decomposition pathways of TATB when exposed to abnormal thermal environments.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Radius measurement via super-resolution microscopy enables the development of a variable radii proximity labeling platform

The elucidation of protein interaction networks is critical to understanding fundamental biology as well as developing new therapeutics. Proximity labeling platforms (PLPs) are state-of-the-art technologies that enable the discovery and delineation of biomolecular networks through the identification of protein-protein interactions. These platforms work via catalytic generation of reactive probes at a biological region of interest; these probes then diffuse through solution and covalently “tag” proximal biomolecules. The physical distance that the probes diffuse determines the effective labeling radius of the PLP and is a critical parameter that influences the scale and resolution of interactome mapping. As such, by expanding the degrees of labeling resolution offered by PLPs, it is possible to better capture the various size scales of interactomes. At present, however, there is little quantitative understanding of the labeling radii of different PLPs. Here, we report the development of a superresolution microscopy-based assay for the direct quantification of PLP labeling radii. Using this assay, we provide direct extracellular measurements of the labeling radii of state-of-the-art antibody-targeted PLPs, including the peroxidase-based phenoxy radical platform (269 ± 41 nm) and the high-resolution iridium-catalyzed µMap technology (54 ± 12 nm). Last, we apply these insights to the development of a molecular diffusion-based approach to tuning PLP resolution and introduce a new aryl-azide-based µMap platform with an intermediate labeling radius (80 ± 28 nm).

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2019

Using flow cytometry and multistage machine learning to discover label-free signatures of algal lipid accumulation

Most applications of flow cytometry or cell sorting rely on the conjugation of fluorescent dyes to specific biomarkers. However, labeled biomarkers are not always available, they can be costly, and they may disrupt natural cell behavior. Label-free quantification based upon machine learning approaches could help correct these issues, but label replacement strategies can be very difficult to discover when applied labels or other modifications in measurements inadvertently modify intrinsic cell properties. Here we demonstrate a new, but simple approach based upon feature selection and linear regression analyses to integrate statistical information collected from both labeled and unlabeled cell populations and to identify models for accurate label-free single-cell quantification. We verify the method’s accuracy to predict lipid content in algal cells (Picochlorum soloecismus) during a nitrogen starvation and lipid accumulation time course. Our general approach is expected to improve label-free single-cell analysis for other organisms or pathways, where biomarkers are inconvenient, expensive, or disruptive to downstream cellular processes.

42 ENGINEERING↗
DOE OSTI2016

Classification with asymmetric label noise: Consistency and maximal denoising

In many real-world classification problems, the labels of training examples are randomly corrupted. Most previous theoretical work on classification with label noise assumes that the two classes are separable, that the label noise is independent of the true class label, or that the noise proportions for each class are known. In this work, we give conditions that are necessary and sufficient for the true class-conditional distributions to be identifiable. These conditions are weaker than those analyzed previously, and allow for the classes to be nonseparable and the noise levels to be asymmetric and unknown. The conditions essentially state that a majority of the observed labels are correct and that the true class-conditional distributions are “mutually irreducible,” a concept we introduce that limits the similarity of the two distributions. For any label noise problem, there is a unique pair of true class-conditional distributions satisfying the proposed conditions, and we argue that this pair corresponds in a certain sense to maximal denoising of the observed distributions. Our results are facilitated by a connection to “mixture proportion estimation,” which is the problem of estimating the maximal proportion of one distribution that is present in another. We establish a novel rate of convergence result for mixture proportion estimation, and apply this to obtain consistency of a discrimination rule based on surrogate loss minimization. Experimental results on benchmark data and a nuclear particle classification problem demonstrate the efficacy of our approach. MSC 2010 subject classifications: Primary 62H30; secondary 68T10. Keywords and phrases: Classification, label noise, mixture proportion estimation, surrogate loss, consistency.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2021

An Open Combinatorial Diffraction Dataset Including Consensus Human and Machine Learning Labels with Quantified Uncertainty for Training New Machine Learning Models

Modern machine learning and autonomous experimentation schemes in materials science rely on accurate analysis of the data ingested by these models. Unfortunately, accurate analysis of the underlying data can be difficult, even for domain experts, complicating the training of the models intended to drive experiments. This is especially true when the goal is to identify the presence of weak signatures in diffraction or spectroscopic datasets. In this work, we examine a set of as-obtained diffraction data that track the phase transition from monoclinic to tetragonal in a Nb-doped VO2 film as a function of temperature and dopant concentration. We then task a set of domain experts and a set of machine learning experts with identifying which phase is present in each diffraction pattern manually and algorithmically, respectively; in both cases, the labels can vary dramatically, especially at the phase boundaries. We use the mode of the labels and the Shannon entropy as a method to capture, preserve and propagate consensus labels and their variance. Further we use the expert labels as a benchmark and demonstrate the use of Shannon entropy weighted scoring to test the performance of machine learning generated labels. Finally, we propose a material data challenge centered around generating improved labeling algorithms. This real-world dataset curated with expert labels can act as test bed for new algorithms. The raw data, annotations and code used in this study are all available online at data.gov and the interested reader is encouraged to replicate and improve the existing models

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2018

Proximity labeling of cis-ligands of CD22/Siglec-2 reveals stepwise α2,6 sialic acid-dependent and -independent interactions

Highlights: • Cis-ligands of CD22 are efficiently labeled by proximity labeling with tyramide. • Cis-ligands in relative proximity of CD22 in the absence of lectin-glycan interaction. • Lectin-glycan interaction further recruits cis-ligands to the close proximity of CD22. • Very low affinity interaction is sufficient for recruitment of cis-ligands to CD22. Lectins expressed on the cell surface are often bound and regulated by the membrane molecules containing the glycan ligands on the same cell (cis-ligands). However, molecular nature and function of cis-ligands are generally poorly understood partly because of weak interaction between lectins and glycan ligands. Cis-ligands are most extensively studied in CD22 (also known as Siglec-2), an inhibitory B lymphocyte receptor specifically recognizing α2,6 sialic acids. CD22, CD45 and IgM are suggested to be ligands of CD22. Here we labeled molecules in the proximity of CD22 in situ on B cell surface using biotin-tyramide. Molecules including CD22, CD45 and IgM were labeled in wild-type but not ST6GalI{sup −/-} B cells that lack α2,6 sialic acids, indicating that these molecules associate with CD22 by lectin-glycan interaction, and are therefore cis-ligands. In ST6GalI{sup −/-} B cells, these cis-ligands are located in a slightly more distance from CD22. Thus, the lectin-glycan interaction recruits cis-ligands already located in the relative proximity of CD22 through non-lectin-glycan interaction to the close proximity. Moreover, cis-ligands are labeled in Cmah{sup −/-} B cells that lack Neu5Gc preferred by mouse CD22 as efficiently as in wild-type B cells, indicating that very low affinity lectin-glycan interaction is sufficient for recruiting cis-ligands, and can be detected by proximity labeling. Thus, proximity labeling with tyramide appears to be a useful method to identify cis-ligands and to analyze their interaction with the lectins.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2018

Incorporation of isotopic, fluorescent, and heavy-atom-modified nucleotides into RNAs by position-selective labeling of RNA

Here, site-specific incorporation of labeled nucleotides is an extremely useful synthetic tool for many structural studies (e.g., NMR, electron paramagnetic resonance (EPR), fluorescence resonance energy transfer (FRET), and X-ray crystallography) of RNA. However, specific-position-labeled RNAs >60 nt are not commercially available on a milligram scale. Position-selective labeling of RNA (PLOR) has been applied to prepare large RNAs labeled at desired positions, and all the required reagents are commercially available. Here, we present a step-by-step protocol for the solid-liquid hybrid phase method PLOR to synthesize 71-nt RNA samples with three different modification applications, containing (i) a (CN)-C-13-N-15-labeled segment; (ii) discrete residues modified with Cy3, Cy5, or biotin; or (iii) two iodo-U residues. The flexible procedure enables a wide range of downstream biophysical analyses using precisely localized functionalized nucleotides. All three RNAs were obtained in <2 d, excluding time for preparing reagents and optimizing experimental conditions. With optimization, the protocol can be applied to other RNAs with various labeling schemes, such as ligation of segmentally labeled fragments.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

Classifying handedness in chiral nanomaterials using label error robust deep learning

Abstract High-throughput scanning electron microscopy (SEM) coupled with classification using neural networks is an ideal method to determine the morphological handedness of large populations of chiral nanoparticles. Automated labeling removes the time-consuming manual labeling of training data, but introduces label error, and subsequently classification error in the trained neural network. Here, we evaluate methods to minimize classification error when training from automated labels of SEM datasets of chiral Tellurium nanoparticles. Using the mirror relationship between images of opposite handed particles, we artificially create populations of varying label error. We analyze the impact of label error rate and training method on the classification error of neural networks on an ideal dataset and on a practical dataset. Of the three training methods considered, we find that a pretraining approach yields the most accurate results across label error rates on ideal datasets, where size and other morphological variables are held constant, but that a co-teaching approach performs the best in practical application.

36 MATERIALS SCIENCE↗
DOE OSTI2025

A machine-learning-driven data labeling pipeline for scientific analysis in MLExchange

This study introduces a novel labeling pipeline to accelerate the labeling process of scientific data sets by using artificial intelligence (AI)-guided tagging techniques. This pipeline includes a set of interconnected web-based graphical user interfaces (GUIs), where Data Clinic and MLCoach enable the preparation of machine learning (ML) models for data reduction and classification, respectively, while Label Maker is used for label assignment. Throughout this pipeline, data can be accessed through a direct connection to a file system or through Tiled for access through Hypertext Transfer Protocol (HTTP). Our experimental results present three use cases where this labeling pipeline has been instrumental for the study of large X-ray scattering data sets in the area of pattern recognition, the remote analysis of resonant soft X-ray scattering data and the fine-tuning process of foundation models. These use cases highlight the labeling capabilities of this pipeline, including the ability to label large data sets in a short period of time, to perform remote data analysis while minimizing data movement and to enhance the fine-tuning process of complex ML models with human involvement.

Chavez, Tanny (ORCID:0000000193172896)↗
DOE OSTI2025

A simple and highly efficient protocol for 13 C-labeling of plant cell wall for structural and quantitative analyses via solid-state nuclear magnetic resonance

Plant cell walls are made of a complex network of interacting polymers that play a critical role in plant development and responses to environmental changes. Thus, improving plant biomass and fitness requires the elucidation of the structural organization of plant cell walls in their native environment. The 13 C-based multi-dimensional solid-state nuclear magnetic resonance (ssNMR) has been instrumental in revealing the structural information of plant cell walls through 2D and 3D correlation spectral analyses. However, the requirement of enriching plants with 13 C limits the applicability of this method. To our knowledge, there is only a very limited set of methods currently available that achieve high levels of 13 C-labeling of plant materials using 13 CO 2 , and most of them require large amounts of 13 CO 2 in larger growth chambers. In this study, a simplified protocol for 13C-labeling of plant materials is introduced that allows ca 60% labeling of the cell walls, as quantified by comparison with commercially labeled samples. This level of 13 C-enrichment is sufficient for all conventional 2D and 3D correlation ssNMR experiments for detailed analysis of plant cell wall structure. The protocol is based on a convenient and easy setup to supply both 13 C-labeled glucose and 13 CO 2 using a vacuum-desiccator. The protocol does not require large amounts of 13 CO 2 . This study shows that our 13 C-labeling of plant materials can make the accessibility to ssNMR technique easy and affordable. The derived high-resolution 2D and 3D correlation spectra are used to extract structural information of plant cell walls. This helps to better understand the influence of polysaccharide-polysaccharide interaction on plant performance and allows for a more precise parametrization of plant cell wall models.

09 BIOMASS FUELS↗
DOE OSTI2022

Tag you're it: Application of stable isotope labeling and LC-MS to identify the precursors of specialized metabolites in plants

Untargeted liquid chromatography/mass spectrometry (LC-MS) can contribute a comprehensive and unbiased picture of the metabolic space of plants. These data can be used to quantify natural metabolite variation for genome wide association studies, to compare global metabolic responses from environmental or genetic perturbations, and to identify previously undescribed metabolites in Nature. A major limitation with untargeted metabolomics is the classification and identification of the thousands of metabolite features that can be detected in a single analytical run. Isotopic labeling improves the informational value of these datasets by categorizing metabolites as being derived from specific upstream precursors and/or to known metabolic pathways. When a 13 C-labeled precursor is fed to either a plant or tissue, the downstream metabolites produced from it have a higher m/z value than the molecules in the pre-existing pool, generating an m/z peak pair that can be specifically identified within the MS data. In this paper, we outline methods and principles to consider when supplementing untargeted MS data with isotopic labeling, including how to choose the appropriate isotopic label, grow and feed plant tissues to maximize label uptake and incorporation into derivatives, optimize LC-MS methods, and interpret the resulting labeling data. Although the focus here is on annotation of amino acid-derived metabolites using LC-MS, we anticipate that the methods are generally adaptable to other precursors, plant species, and chromatographic approaches.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2017

Immunophotoaffinity labeling of binders of 1-methyladenine, the oocyte maturation-inducing hormone of starfish

Starfish oocytes are arrested at the prophase stage of the first meiotic division in the ovary and resume meiosis by the stimulus of 1-methyladenine (1-MeAde), the oocyte maturation-inducing hormone of starfish. Putative 1-MeAde receptors on the oocyte surface have been suggested, but not yet been biochemically characterized. Immunophotoaffinity labeling, i.e., photoaffinity labeling combined with immunochemical detection, was attempted to detect unknown 1-MeAde binders including putative maturation-inducing hormone receptors in starfish oocytes. When the oocyte crude membrane fraction or its Triton X-100/EDTA extract was incubated with N{sup 6}-[6-(5-azido-2-nitrobenzoyl)aminohexyl]carboxamidomethyl-1-methyladenine and then photo-irradiated, followed by western blotting with antibody that was raised against a 1-MeAde hapten, a single band with M{sub r} of 47.5 K was detected. The band was lost when extract was heated at 100 °C. A similar 47.5 K band was detected in the crude membrane fraction of testis as well. Upon labeling with whole cells, this band was detected in immature and maturing oocytes, but only faintly in mature oocytes. As judged from these results, this 1-MeAde binder might be a possible candidate of the starfish maturation-inducing hormone receptors. - Highlights: • Synthesis of photoaffinity labeling reagents for 1-methyladenine binders of starfish. • Immunochemical detection of photoaffinity-labeled 1-methyladenine binders. • Immunophotoaffinity labeling of a 47.5 K 1-methyladenine binder in oocytes and testis. • A possible candidate of oocyte maturation-inducing hormone receptors of starfish.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2024

Endogenous Production and Vibrational Analysis of Heavy-Isotope-Labeled Peptides from Cyanobacteria

Stable isotope labeling is an extremely useful tool for characterizing the structure, tracing the metabolism, and imaging the distribution of natural products in living organisms using mass-sensitive measurement techniques. In this study, a cyanobacterium was cultured in 15 N/ 13 C-enriched media to endogenously produce labeled, bioactive oligopeptides. The extent of heavy isotope incorporation in these peptides was determined with LC–MS, while the overall extent of heavy isotope incorporation in whole cells was studied with nanoSIMS and AFM-IR. Up to 98 % heavy isotope incorporation was observed in labeled cells. Three of the most abundant peptides, microcystin-LR (MCLR), cyanopeptolin-A (CYPA), and aerucyclamide-A (ACAA), were isolated and further studied with Raman and FTIR spectroscopies and DFT calculations. This revealed several IR and Raman active vibrations associated with functional groups not common in ribosomal peptides, like diene, ester, thiazole, thiazoline, and oxazoline groups, which could be suitable for future vibrational imaging studies. More broadly, this study outlines a simple and relatively inexpensive method for producing heavy-labeled natural products. Finally, manipulating the bacterial culture conditions by the addition of specific types and amounts of heavy-labeled nutrients provides an efficient means of producing heavy-labeled natural products for mass-sensitive imaging studies.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2022

Phosphorothioate-Based Site-Specific Labeling of Large RNAs for Structural and Dynamic Studies

Pulsed electron-electron double resonance (PELDOR) spectroscopy, X-ray scattering interferometry (XSI), and single-molecule Forster resonance energy transfer (smFRET) are molecular rulers that provide inter- or intramolecular pair-wise distance distributions in the nanometer range, thus being ideally suitable for structural and dynamic studies of biomolecules including RNAs. The prerequisite for such applications requires site-specific labeling of biomolecules with spin labels, gold nanoparticles, and fluorescent tags, respectively. Recently, site-specific labeling of large RNAs has been achieved by a combination of transcription of an expanded genetic alphabet containing A-T/G-C base pairs and NaM-TPT3 unnatural base pair (UBP) with posttranscriptional modifications at UBP bases by click chemistry or amine-NHS ester reactions. However, due to the bulky sizes of functional groups or labeling probes used, such strategies might cause structural perturbation and decrease the accuracy of distance measurements. Here, we synthesize an a-thiophosphorylated variant of rTPT3TP (rTPT3aS), which allows for post-transcriptional site-specific labeling of large RNAs at the internal a-phosphate backbone via maleimide-modified probes. Subsequent PELDOR, XSI, and smFRET measurements result in narrower distance distributions than labeling at the TPT3 base. The presented strategy provides a new route to empower the molecular rulers for structural and dynamic studies of large RNA and its complex.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2018

Automated Labeling of Electron Microscopy Images Using Deep Learning

© 2018 IEEE. Searching for scientific data requires metadata providing a relevant context. Today, generating metadata is a time and labor intensive manual process that is often neglected, and important datasets are not accessible through search. We investigate the use of machine learning to generalize metadata from a subset of labeled data, thus increasing the availability of meaningful metadata for search. Specifically, we consider electron microscopy images collected at the National Center for Electron Microscopy (NCEM) at the Lawrence Berkeley National Laboratory (LBNL) and use deep learning to discern characteristics from a small subset of labeled images and transfer labels to the entire image corpus. Relatively small training set sizes and a minimum resolution of 512\times 512 pixels required by the application domain pose unique challenges. We overcome these challenges by using a simple yet powerful convolutional network architecture that limits the number of free parameters to lower the required amount of computational power and reduce the risk of overfitting. We achieve a classification accuracy of approximately 80% in discerning between images recorded in two operating modes of the electron microscope-transmission electron microscopy (TEM) and scanning transmission electron microscopy (STEM). We use transfer learning-i.e., re-using the pre-trained convolution layers from the TEM vs. STEM classification problem-to generalize labels and achieve an accuracy of approximately 70 % despite current experiments being limited to small training data sets. We present these predictions as suggestions to domain scientists through a web-based tool to accelerate the labeling process with the goal of further validating our approach and improving the accuracy of automatically created labels.

Weber, Gunther H↗