pith. sign in

arxiv: 2606.10352 · v2 · pith:TJBOXQMVnew · submitted 2026-06-09 · ❄️ cond-mat.mtrl-sci

Neural electron backscatter diffraction

I-Tzu Huang , Marat I. Latypov This is my paper

Pith reviewed 2026-06-27 12:58 UTC · model grok-4.3

classification ❄️ cond-mat.mtrl-sci
keywords electron backscatter diffractionKikuchi patternsneural networksmicrostructure analysisdata compressioncontinuous representationsuper-resolution
0
0 comments X

The pith

A coordinate-based neural network represents EBSD scans as continuous four-dimensional fields of Kikuchi diffraction intensity.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper introduces neural EBSD to treat discrete grid scans as continuous differentiable fields in the specimen-detector domain. It compares a joint neural mapping of all four coordinates to intensity against a factorized version that separates specimen-domain coefficient fields from detector-domain basis patterns. On recrystallized and additively manufactured Inconel 718, the factorized formulation reconstructs patterns with map-averaged errors below 1 percent of maximum intensity. This approach yields super-resolution, off-grid querying, derivative-based boundary localization, and 700-fold data compression while retaining full-pattern access.

Core claim

We introduce neural EBSD, which treats an EBSD scan as a continuous, differentiable four-dimensional field of Kikuchi diffraction intensity (in specimen--detector domain) and then represents it with a coordinate-based neural network. We develop and compare two formulations: a joint formulation that maps all four coordinates to intensity, and a factorized formulation that combines continuous specimen-domain coefficient fields with learned detector-domain basis patterns. Tested on recrystallized and additively manufactured Inconel 718, the factorized formulation shows better accuracy in reconstructing Kikuchi patterns that have map-averaged errors below 1% of the maximum intensity.

What carries the argument

Factorized coordinate-based neural network separating continuous specimen-domain coefficient fields from learned detector-domain basis patterns to represent the four-dimensional Kikuchi intensity field.

If this is right

  • Full-pattern super-resolution in the specimen frame
  • Continuous querying along arbitrary off-grid paths
  • Spatially continuous boundary and heterogeneity localization from analytical spatial derivatives
  • 700-fold compression by storing network weights and learned bases instead of raw patterns

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The continuous differentiable representation could couple directly with deformation simulations that require smooth fields rather than gridded data.
  • Off-grid querying opens the possibility of adaptive, non-uniform EBSD scanning strategies that concentrate measurements where heterogeneity is highest.
  • The compression and on-demand access approach could apply to other large diffraction or imaging datasets in materials science where grid storage becomes prohibitive.

Load-bearing premise

Kikuchi diffraction intensity varies sufficiently smoothly in the specimen-detector domain and factorizes into continuous specimen-domain coefficient fields plus learned detector-domain basis patterns to allow accurate neural representation without loss of orientation or dislocation information.

What would settle it

Reconstruction errors exceeding 1% of maximum intensity on the Inconel 718 samples, or loss of fidelity when the network is queried along continuous off-grid paths, would show the representation fails to capture the underlying field.

Figures

Figures reproduced from arXiv: 2606.10352 by I-Tzu Huang, Marat I. Latypov.

Figure 1
Figure 1. Figure 1: Overview of the present study. tensity as a continuous function of all four coordi￾nates (x, y, u, v); and (ii) a factorized formulation (TF-INR), in which a SIREN-type network learns continuous coefficient fields in the specimen domain that linearly combine a set of learnable basis pat￾terns in the detector domain (Figure 1c). With a case study on Ni-base superalloys, we find that the factorized formulati… view at source ↗
Figure 2
Figure 2. Figure 2: Reconstructions of the RX718 map with 4D-INR and TF-INR compared in terms of (a) orientation deviation maps [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Reconstruction of EBSD data evaluated in terms of (a) orientation maps obtained with Hough indexing of measured [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Learned TF-INR representation: (a) coefficient field [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Demonstration of super-resolution: (a) hold-out of every other pixel (red cross) during training to emulate coarse-grid [PITH_FULL_IMAGE:figures/full_fig_p009_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Demonstration of sub-grid path sampling: (a-b) sub-grid query between EBSD pixels in two directions in RX718 and [PITH_FULL_IMAGE:figures/full_fig_p010_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Grain boundary analysis for RX718 and AM718 microstructures: (a) conventional disorientation-based segmentation [PITH_FULL_IMAGE:figures/full_fig_p011_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: (a) Explained variance of randomly sampled 200 000 patterns from the two EBSD datasets; (b) visual illustration of [PITH_FULL_IMAGE:figures/full_fig_p014_8.png] view at source ↗
read the original abstract

In a polycrystalline microstructure, orientation and dislocation content vary smoothly within grains, and the grain boundaries between them are continuous curves. Electron backscatter diffraction (EBSD) records this continuum on a discrete grid with every subsequent analysis (from indexing to advanced pattern processing) confined to that grid. We introduce neural EBSD, which treats an EBSD scan as a continuous, differentiable four-dimensional field of Kikuchi diffraction intensity (in specimen--detector domain) and then represents it with a coordinate-based neural network. We develop and compare two formulations: a joint formulation that maps all four coordinates to intensity, and a factorized formulation that combines continuous specimen-domain coefficient fields with learned detector-domain basis patterns. Tested on recrystallized and additively manufactured Inconel 718, the factorized formulation shows better accuracy in reconstructing Kikuchi patterns that have map-averaged errors below 1% of the maximum intensity. Beyond reconstruction, it provides full-pattern super-resolution in the specimen frame, continuous querying along arbitrary off-grid paths, as well as spatially continuous boundary and heterogeneity localization from analytical spatial derivatives. Storing the network weights and learned bases in place of the raw patterns in a large dataset offers a 700-fold compression while preserving on-demand access to the full patterns for downstream analyses.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper introduces neural EBSD, representing EBSD scans as a continuous, differentiable 4D Kikuchi intensity field via coordinate-based neural networks. It compares a joint formulation (mapping all four coordinates to intensity) against a factorized formulation (specimen-domain coefficient fields multiplied by learned detector-domain bases). On recrystallized and additively manufactured Inconel 718, the factorized version achieves map-averaged reconstruction errors below 1% of peak intensity, while enabling full-pattern super-resolution in the specimen frame, continuous off-grid querying, analytical spatial derivatives for boundary/heterogeneity localization, and 700-fold compression via stored weights and bases.

Significance. If the reconstruction accuracy and preservation of downstream information hold, the work provides a practical continuous representation for EBSD data that could reduce storage demands in large datasets while enabling new analyses such as derivative-based boundary detection. The empirical testing on two distinct Inconel 718 microstructures and the explicit comparison of the two formulations are strengths; the compression factor and on-demand pattern access are concrete advantages for practical adoption.

major comments (2)
  1. [Abstract / Results] Abstract and results section: The central claim that the representation 'preserves on-demand access to the full patterns for downstream analyses' rests on map-averaged reconstruction error <1% of maximum intensity, but this metric does not establish that orientation indexing or GND density maps remain accurate. Grain boundaries and dislocations violate the smoothness premise of the factorized formulation (continuous coefficient fields imes fixed bases), and averaging can mask localized errors at those sites; a direct comparison of indexed orientations or GND densities computed from the neural patterns versus the original grid is required to support the claim.
  2. [Methods] Methods / factorized formulation: The assumption that 4D Kikuchi intensity admits a low-rank separable decomposition into specimen-domain coefficients and detector-domain bases is load-bearing for the reported accuracy advantage of the factorized model, yet no quantitative assessment (e.g., singular-value spectrum of the data tensor or sensitivity to rank choice) is supplied to show that the decomposition does not discard orientation or dislocation information.
minor comments (2)
  1. [Abstract] The abstract states '700-fold compression' without specifying the baseline (raw pattern storage size, bit depth, or whether it includes the network weights plus bases); a precise definition and comparison table would clarify the claim.
  2. [Results] No error bars, dataset sizes, or cross-validation procedure are mentioned for the <1% error figure; adding these in the results section would strengthen reproducibility.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. We address each major comment below, agreeing that the current evidence for downstream preservation is indirect and that additional quantitative support for the separability assumption would strengthen the manuscript.

read point-by-point responses

  1. Referee: [Abstract / Results] Abstract and results section: The central claim that the representation 'preserves on-demand access to the full patterns for downstream analyses' rests on map-averaged reconstruction error <1% of maximum intensity, but this metric does not establish that orientation indexing or GND density maps remain accurate. Grain boundaries and dislocations violate the smoothness premise of the factorized formulation (continuous coefficient fields imes fixed bases), and averaging can mask localized errors at those sites; a direct comparison of indexed orientations or GND densities computed from the neural patterns versus the original grid is required to support the claim.

    Authors: We agree that map-averaged reconstruction error alone does not fully establish preservation of derived quantities such as indexed orientations or GND densities, particularly near grain boundaries where the smoothness assumption is most stressed. The manuscript demonstrates sub-1% error on both recrystallized and AM Inconel 718 (which contain boundaries and defects) and shows that the factorized model enables analytical derivatives for boundary localization, but we acknowledge the need for explicit validation. We will add direct comparisons of orientation indexing accuracy and GND density maps computed from neural-reconstructed patterns versus the original data in the revised results section. revision: yes

  2. Referee: [Methods] Methods / factorized formulation: The assumption that 4D Kikuchi intensity admits a low-rank separable decomposition into specimen-domain coefficients and detector-domain bases is load-bearing for the reported accuracy advantage of the factorized model, yet no quantitative assessment (e.g., singular-value spectrum of the data tensor or sensitivity to rank choice) is supplied to show that the decomposition does not discard orientation or dislocation information.

    Authors: The factorized formulation is motivated by the physical separation between specimen coordinates (encoding orientation and defect variation) and detector coordinates (fixed diffraction geometry), and its superior reconstruction accuracy relative to the joint model provides empirical support that essential information is retained. However, we did not include a singular-value spectrum of the 4D data tensor or a rank-sensitivity study. We will add both analyses to the methods section in revision to quantify the effective rank and confirm that orientation/dislocation content is not discarded. revision: yes

Circularity Check

0 steps flagged

No circularity; empirical performance claims are independent

full rationale

The paper introduces coordinate-based neural representations (joint and factorized) for 4D Kikuchi intensity fields in EBSD data. Central claims rest on measured reconstruction errors (<1% map-averaged) and derived capabilities (super-resolution, continuous querying, compression) evaluated directly against input patterns from Inconel 718 samples. No equations or steps reduce by construction to inputs, no self-citations are load-bearing, and no fitted parameters are relabeled as predictions. The smoothness premise is an explicit modeling assumption whose validity is checked empirically rather than assumed tautologically. The derivation chain is self-contained against external data benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the domain assumption that diffraction intensity admits a smooth neural representation; no free parameters or invented entities are explicitly introduced in the abstract.

axioms (1)
  • domain assumption EBSD intensity varies smoothly within grains and can be represented as a continuous differentiable 4D field in specimen-detector coordinates.
    Invoked when the paper states that orientation and dislocation content vary smoothly and that the scan can be treated as a continuous field.

pith-pipeline@v0.9.1-grok · 5749 in / 1340 out tokens · 21507 ms · 2026-06-27T12:58:00.709486+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

56 extracted references · 2 canonical work pages

  1. [1]

    A. J. Schwartz, M. Kumar, B. L. Adams, D. P. Field, Electron backscatter diffraction in materials science, Vol. 2, Springer, 2009

    2009

  2. [2]

    A. J. Wilkinson, T. B. Britton, Strains, planes, and EBSD in materials science, Materials today 15 (9) (2012) 366–376

    2012

  3. [3]

    Kunze, S

    K. Kunze, S. Wright, B. L. Adams, D. J. Dingley, Ad- vances in automatic ebsp single orientation measure- ments, Texture, Stress, and Microstructure 20 (1-4) (1993) 41–54

    1993

  4. [4]

    Krieger Lassen, Automatic localisation of electron backscattering pattern bands from hough transform, Materials Science and Technology 12 (10) (1996) 837– 843

    N. Krieger Lassen, Automatic localisation of electron backscattering pattern bands from hough transform, Materials Science and Technology 12 (10) (1996) 837– 843

    1996

  5. [5]

    Y. H. Chen, S. U. Park, D. Wei, G. Newstadt, M. A. Jackson, J. P. Simmons, M. De Graef, A. O. Hero, A dictionary approach to electron backscatter diffraction indexing, Microscopy and Microanalysis 21 (3) (2015) 739–752

    2015

  6. [6]

    Bachmann, R

    F. Bachmann, R. Hielscher, H. Schaeben, Grain de- tection from 2D and 3D EBSD data—specification of the MTEX algorithm, Ultramicroscopy 111 (12) (2011) 1720–1733

    2011

  7. [7]

    Pantleon, Resolving the geometrically necessary dis- location content by conventional electron backscatter- ing diffraction, Scripta Materialia 58 (11) (2008) 994– 997

    W. Pantleon, Resolving the geometrically necessary dis- location content by conventional electron backscatter- ing diffraction, Scripta Materialia 58 (11) (2008) 994– 997

    2008

  8. [8]

    S. I. Wright, M. M. Nowell, D. P. Field, A review of strain analysis using electron backscatter diffraction, Microscopy and microanalysis 17 (3) (2011) 316–329

    2011

  9. [9]

    Kamaya, Assessment of local deformation using EBSD: Quantification of local damage at grain bound- aries, Materials characterization 66 (2012) 56–67

    M. Kamaya, Assessment of local deformation using EBSD: Quantification of local damage at grain bound- aries, Materials characterization 66 (2012) 56–67

    2012

  10. [10]

    D. M. Saylor, B. S. El-Dasher, B. L. Adams, G. S. Rohrer, Measuring the five-parameter grain-boundary distribution from observations of planar sections, Met- allurgical and Materials Transactions A 35 (7) (2004) 1981–1989. 15

    2004

  11. [11]

    Wang, J.-C

    F. Wang, J.-C. Stinville, M. Charpagne, M. P. Ech- lin, S. R. Agnew, T. M. Pollock, M. De Graef, D. S. Gianola, Dislocation cells in additively manufactured metallic alloys characterized by electron backscatter diffraction pattern sharpness, Materials Characteriza- tion 197 (2023) 112673

    2023

  12. [12]

    Thome, S

    P. Thome, S. Medghalchi, J. Frenzel, J. Schreuer, G. Eggeler, Ni-base superalloy single crystal (sx) mo- saicity characterized by the rotation vector base line electron back scatter diffraction (rvb-ebsd) method, Ul- tramicroscopy 206 (2019) 112817

    2019

  13. [13]

    L. N. Brewer, P. G. Kotula, J. R. Michael, Multivariate statistical approach to electron backscattered diffrac- tion, Ultramicroscopy 108 (6) (2008) 567–578

    2008

  14. [14]

    A. J. Wilkinson, D. M. Collins, Y. Zayachuk, R. Ko- rla, A. Vilalta-Clemente, Applications of multivariate statistical methods and simulation libraries to anal- ysis of electron backscatter diffraction and transmis- sion kikuchi diffraction datasets, Ultramicroscopy 196 (2019) 88–98

    2019

  15. [15]

    T. P. McAuliffe, D. Dye, T. B. Britton, Spherical- angular dark field imaging and sensitive microstructural phase clustering with unsupervised machine learning, Ultramicroscopy 219 (2020) 113132

    2020

  16. [16]

    Chauniyal, P

    A. Chauniyal, P. Thome, M. Stricker, Employing constrained nonnegative matrix factorization for mi- crostructure segmentation, Microscopy and Microanal- ysis 30 (4) (2024) 712–723

    2024

  17. [17]

    Z. T. Varley, G. S. Rohrer, M. De Graef, Accelerating dictionary indexing of electron backscatter diffraction patterns with pca and quantization, Scientific Reports 16 (2026) 4382

    2026

  18. [18]

    Calvat, C

    M. Calvat, C. Bean, D. Anjaria, H. Park, H. Wang, K. Vecchio, J. Stinville, Learning metal microstructural heterogeneity through spatial mapping of diffraction la- tent space features, npj Computational Materials 11 (1) (2025) 284

    2025

  19. [19]

    Liu, C.-K

    Y.-C. Liu, C.-K. Yeh, S.-P. Tsai, P.-Y. Tung, Learning crystallographic orientations from electron backscatter diffraction patterns using variational autoencoder, Cell Reports Physical Science 6 (10)

  20. [20]

    J. Wang, M. Calvat, J. C. Stinville, M. I. Latypov, Lattice genome: representation and analysis of hetero- geneous crystalline microstructures, arXiv:2606.09611 (2026)

    Pith/arXiv arXiv 2026

  21. [21]

    Calvat, D

    M. Calvat, D. Anjaria, H. Wang, K. Vecchio, J.-C. Stinville, Kikuchi pattern dataset from wrought and as- built additively manufactured superalloys (Sep. 2025). doi:10.5061/DRYAD.ZCRJDFNR9

    work page doi:10.5061/dryad.zcrjdfnr9 2025

  22. [22]

    M. I. Latypov, M. K¨ uhbach, I. J. Beyerlein, J.-C. Stinville, L. S. Toth, T. M. Pollock, S. R. Kalidindi, Application of chord length distributions and principal component analysis for quantification and representa- tion of diverse polycrystalline microstructures, Materi- als Characterization 145 (2018) 671–685

    2018

  23. [23]

    Mildenhall, P

    B. Mildenhall, P. P. Srinivasan, M. Tancik, J. T. Bar- ron, R. Ramamoorthi, R. Ng, Nerf: Representing scenes as neural radiance fields for view synthesis, Communi- cations of the ACM 65 (1) (2021) 99–106

    2021

  24. [24]

    J. J. Park, P. Florence, J. Straub, R. Newcombe, S. Lovegrove, Deepsdf: Learning continuous signed dis- tance functions for shape representation, in: Proceed- ings of the IEEE/CVF conference on computer vision and pattern recognition, 2019, pp. 165–174

    2019

  25. [25]

    Sitzmann, J

    V. Sitzmann, J. Martel, A. Bergman, D. Lindell, G. Wetzstein, Implicit neural representations with peri- odic activation functions, Advances in neural informa- tion processing systems 33 (2020) 7462–7473

    2020

  26. [26]

    Str¨ umpler, J

    Y. Str¨ umpler, J. Postels, R. Yang, L. Van Gool, F. Tombari, Implicit neural representations for image compression, in: Computer Vision – ECCV 2022: 17th European Conference, Tel Aviv, Israel, October 23–27, 2022, Proceedings, Part XXVI, Springer-Verlag, 2022, pp. 74–91

    2022

  27. [27]

    Pezzoli, F

    M. Pezzoli, F. Antonacci, A. Sarti, Implicit neural rep- resentation with physics-informed neural networks for the reconstruction of the early part of room impulse responses, in: Proceedings of the 10th Convention of the European Acoustics Association Forum Acusticum 2023, European Acoustics Association, 2023, pp. 2177– 2184

    2023

  28. [28]

    P. Xie, Y. Liu, C. Liu, C. Song, X. Zhang, Simulta- neous suppression of seismic random and erratic noise using PINN with high-frequency preservation, Journal of Geophysics and Engineering 22 (6) (2025) 1796–1808

    2025

  29. [29]

    Mildenhall, P

    B. Mildenhall, P. Hedman, R. Martin-Brualla, P. P. Srinivasan, J. T. Barron, Nerf in the dark: High dy- namic range view synthesis from noisy raw images, in: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2022, pp. 16190–16199

    2022

  30. [30]

    J. Xu, D. Moyer, B. Gagoski, J. E. Iglesias, P. E. Grant, P. Golland, E. Adalsteinsson, Nesvor: implicit neural representation for slice-to-volume reconstruction in mri, IEEE transactions on medical imaging 42 (6) (2023) 1707–1719

    2023

  31. [31]

    S. Ye, L. Shen, M. T. Islam, L. Xing, Super-resolution biomedical imaging via reference-free statistical implicit neural representation, Physics in Medicine & Biology 68 (20) (2023) 205020

    2023

  32. [32]

    Molaei, A

    A. Molaei, A. Aminimehr, A. Tavakoli, A. Kazerouni, B. Azad, R. Azad, D. Merhof, Implicit neural repre- sentation in medical imaging: A comparative survey (2023).arXiv:2307.16142

    arXiv 2023

  33. [33]

    Moussaoui, D

    Y. Moussaoui, D. Mateus, N. Taheri, S. Moussaoui, S. Carlier, Thomas an Stute, Implicit neural repre- sentations for end-to-end pet reconstruction, in: 2025 IEEE 22nd International Symposium on Biomedical Imaging (ISBI), 2025, pp. 1–5, hAL Id: hal-04999198. doi:10.1109/ISBI60581.2025.10980961

    work page doi:10.1109/isbi60581.2025.10980961 2025

  34. [34]

    C.-M. Hsu, S. Xu, M. J. Lynch, Y. Wang, X. Liu, C. Williams, X. Wang, L. Qian, K. G. Field, X. Qian, INR-TEM: Robust cavity detection in multifocus tem images via implicit neural representations, APL Ma- chine Learning 4 (1)

  35. [35]

    A. Chen, Z. Xu, A. Geiger, J. Yu, H. Su, TensoRF: Tensorial radiance fields, in: Computer Vision – ECCV 2022, Springer, 2022, pp. 333–350

    2022

  36. [36]

    Y. Xie, T. Takikawa, S. Saito, O. Litany, S. Yan, N. Khan, F. Tombari, J. Tompkin, V. Sitzmann, S. Sridhar, Neural fields in visual computing and be- yond, Computer Graphics Forum 41 (2) (2022) 641– 676

    2022

  37. [37]

    Y. Chen, S. Liu, X. Wang, Learning continuous image representation with local implicit image function, in: 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2021, pp. 8624–8634

    2021

  38. [38]

    Mescheder, M

    L. Mescheder, M. Oechsle, M. Niemeyer, S. Nowozin, A. Geiger, Occupancy networks: Learning 3D recon- struction in function space, in: 2019 IEEE/CVF Con- ference on Computer Vision and Pattern Recognition 16 (CVPR), 2019, pp. 4455–4465

    2019

  39. [39]

    M¨ uller, A

    T. M¨ uller, A. Evans, C. Schied, A. Keller, Instant neu- ral graphics primitives with a multiresolution hash en- coding, ACM transactions on graphics (TOG) 41 (4) (2022) 1–15

    2022

  40. [40]

    H. W. ˚Anes, L. A. H. Lervik, O. Natlandsmyr, T. Bergh, E. Prestat, A. Gerlt, A. V. Bugten, E. M. Østvold, Z. Xu, C. Francis, M. Nord, pyxem/kikuchipy: kikuchipy 0.11.5 (2026)

    2026

  41. [41]

    S. I. Wright, B. L. Adams, Automatic analysis of electron backscatter diffraction patterns, Metallurgical Transactions A 23 (1992) 759–767

    1992

  42. [42]

    Bachmann, R

    F. Bachmann, R. Hielscher, H. Schaeben, Texture anal- ysis with MTEX–free and open source software toolbox, Solid state phenomena 160 (2010) 63–68

    2010

  43. [43]

    Krakow, R

    R. Krakow, R. J. Bennett, D. N. Johnstone, Z. Vuk- manovic, W. Solano-Alvarez, S. J. Lain´ e, J. F. Einsle, P. A. Midgley, C. M. F. Rae, R. Hielscher, On three- dimensional misorientation spaces, Proceedings of the Royal Society A: Mathematical, Physical and Engineer- ing Sciences 473 (2206) (2017) 20170274

    2017

  44. [44]

    Z. Ding, E. Pascal, M. De Graef, Indexing of electron back-scatter diffraction patterns using a convolutional neural network, Acta Materialia 199 (2020) 370–382

    2020

  45. [45]

    Q. Shi, Y. Zhou, H. Zhong, D. Loisnard, C. Dan, F. Zhang, Z. Chen, H. Wang, S. Roux, Indexation of electron diffraction patterns at grain boundaries, Mate- rials Characterization 182 (2021) 111553

    2021

  46. [46]

    V. Tong, J. Jiang, A. J. Wilkinson, T. B. Britton, The effect of pattern overlap on the accuracy of high res- olution electron backscatter diffraction measurements, Ultramicroscopy 155 (2015) 62–73

    2015

  47. [47]

    Marquardt, M

    K. Marquardt, M. De Graef, S. Singh, H. Mar- quardt, A. Rosenthal, S. Koizuimi, Quantitative elec- tron backscatter diffraction (ebsd) data analyses us- ing the dictionary indexing (di) approach: Overcoming indexing difficulties on geological materials, American Mineralogist 102 (9) (2017) 1843–1855

    2017

  48. [48]

    Patala, C

    S. Patala, C. A. Schuh, A continuous and one-to-one coloring scheme for misorientations, Acta materialia 59 (2) (2011) 554–562

    2011

  49. [49]

    M. I. Latypov, J.-C. Stinville, J. R. Mayeur, J. M. He- stroffer, T. M. Pollock, I. J. Beyerlein, Insight into microstructure-sensitive elastic strain concentrations from integrated computational modeling and digital im- age correlation, Scripta Materialia 192 (2021) 78–82

    2021

  50. [50]

    Stinville, W

    J.-C. Stinville, W. C. Lenthe, M. P. Echlin, P. G. Calla- han, D. Texier, T. M. Pollock, Microstructural statistics for fatigue crack initiation in polycrystalline nickel-base superalloys, International Journal of Fracture 208 (1) (2017) 221–240

    2017

  51. [51]

    Fotos, A

    G. Fotos, A. Campbell, P. Murray, E. Yakushina, Deep learning enhanced watershed for microstructural analy- sis using a boundary class semantic segmentation, Jour- nal of Materials Science 58 (2023) 14390–14410

    2023

  52. [52]

    Beucher, The watershed transformation applied to image segmentation, Scanning Microscopy 1992 (6)

    S. Beucher, The watershed transformation applied to image segmentation, Scanning Microscopy 1992 (6)

    1992

  53. [53]

    Zaefferer, On the formation mechanisms, spatial res- olution and intensity of backscatter Kikuchi patterns, Ultramicroscopy 107 (2–3) (2007) 254–266

    S. Zaefferer, On the formation mechanisms, spatial res- olution and intensity of backscatter Kikuchi patterns, Ultramicroscopy 107 (2–3) (2007) 254–266

    2007

  54. [54]

    F. J. Humphreys, Review grain and subgrain charac- terisation by electron backscatter diffraction, Journal of Materials Science 36 (2001) 3833–3854

    2001

  55. [55]

    P. W. Trimby, Orientation mapping of nanostruc- tured materials using transmission kikuchi diffraction in the scanning electron microscope, Ultramicroscopy 120 (2012) 16–24

    2012

  56. [56]

    R. R. Keller, R. H. Geiss, Transmission ebsd from 10 nm domains in a scanning electron microscope, Journal of Microscopy 245 (3) (2012) 245–251. 17

    2012