Skip to content

Commit

Permalink
build based on 96305f8
Browse files Browse the repository at this point in the history
  • Loading branch information
Documenter.jl committed Jul 10, 2024
1 parent 74cbf1d commit 030e3b2
Show file tree
Hide file tree
Showing 3 changed files with 3 additions and 3 deletions.
2 changes: 1 addition & 1 deletion dev/.documenter-siteinfo.json
View file
Original file line number Diff line number Diff line change
@@ -1 +1 @@
{"documenter":{"julia_version":"1.10.2","generation_timestamp":"2024-04-24T14:55:31","documenter_version":"1.4.0"}}
{"documenter":{"julia_version":"1.10.4","generation_timestamp":"2024-07-10T07:42:55","documenter_version":"1.4.0"}}
2 changes: 1 addition & 1 deletion dev/index.html
View file
Original file line number Diff line number Diff line change
@@ -1,2 +1,2 @@
<!DOCTYPE html>
<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>PoseErrors.jl · PoseErrors</title><meta name="title" content="PoseErrors.jl · PoseErrors"/><meta property="og:title" content="PoseErrors.jl · PoseErrors"/><meta property="twitter:title" content="PoseErrors.jl · PoseErrors"/><meta name="description" content="Documentation for PoseErrors."/><meta property="og:description" content="Documentation for PoseErrors."/><meta property="twitter:description" content="Documentation for PoseErrors."/><script data-outdated-warner src="assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="assets/documenter.js"></script><script src="search_index.js"></script><script src="siteinfo.js"></script><script src="../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="assets/themeswap.js"></script></head><body><div id="documenter"><nav class="docs-sidebar"><div class="docs-package-name"><span class="docs-autofit"><a href>PoseErrors</a></span></div><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li class="is-active"><a class="tocitem" href>PoseErrors.jl</a><ul class="internal"><li class="toplevel"><a class="tocitem" href="#Setup-for-BOP-dataset-evaluation"><span>Setup for BOP dataset evaluation</span></a></li><li><a class="tocitem" href="#Loading-the-BOP-datasets"><span>Loading the BOP datasets</span></a></li><li class="toplevel"><a class="tocitem" href="#Maximum-Distance-of-Model-Points-for-indistinguishable-views-(MDD-S)"><span>Maximum Distance of Model Points for indistinguishable views (MDD-S)</span></a></li></ul></li><li><a class="tocitem" href="methods/">Methods</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li class="is-active"><a href>PoseErrors.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>PoseErrors.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/rwth-irt/PoseErrors.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/rwth-irt/PoseErrors.jl/blob/main/docs/src/index.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="PoseErrors.jl"><a class="docs-heading-anchor" href="#PoseErrors.jl">PoseErrors.jl</a><a id="PoseErrors.jl-1"></a><a class="docs-heading-anchor-permalink" href="#PoseErrors.jl" title="Permalink"></a></h1><p>A good overview and rationale behind 6D pose error metrics can be found in the <a href="https://bop.felk.cvut.cz/challenges/bop-challenge-2019/#evaluationmethodology">BOP-challenge</a>. They prefer Maximum Symmetry-Aware Surface Distance (MSSD) over Averade Average Distance of Model Points with indistinguishable views (ADD-S = ADI), because they can yield low errors with bad visual alignment. Moreover, ADD-S is dominated by higher-frequency surface parts (e.g. a Thread). Maximum distances do not suffer from the surface sampling density as much.</p><p>However, annotating the symmetries is tedious and heavily depends on the choice of coordinate frames. For a large set of objects like surgical instruments which are not exported in a standardized / symmetry aligned frame, this is impractical. So, similar to (<a href="https://doi.org/10.3390/jimaging8030053">Gorschlüter et al. 2022</a>) we use ADD-S and VDS (<a href="https://doi.org/10.1007/978-3-319-49409-8_52">Hodan et. al 2016</a>) as metrics.</p><h1 id="Setup-for-BOP-dataset-evaluation"><a class="docs-heading-anchor" href="#Setup-for-BOP-dataset-evaluation">Setup for BOP dataset evaluation</a><a id="Setup-for-BOP-dataset-evaluation-1"></a><a class="docs-heading-anchor-permalink" href="#Setup-for-BOP-dataset-evaluation" title="Permalink"></a></h1><p>Extract the BOP datasets as described on their <a href="https://bop.felk.cvut.cz/datasets/">website</a>. Move the detections JSON to the matching datasets test directory and rename it to <code>default_detections.json</code>, e.g. <em>datasets/tless/default_detections.json</em>.</p><p>You could also you the keyword argument <code>detections_file</code> of <code>scene_test_targets</code> to specify another file in the test directory.</p><h2 id="Loading-the-BOP-datasets"><a class="docs-heading-anchor" href="#Loading-the-BOP-datasets">Loading the BOP datasets</a><a id="Loading-the-BOP-datasets-1"></a><a class="docs-heading-anchor-permalink" href="#Loading-the-BOP-datasets" title="Permalink"></a></h2><p>Your first entry points should be the methods to load the according targets. These load the required image files, mesh files, camera parameters, etc.</p><ul><li><code>gt_targets</code>: Loads the ground truth pose as <code>:gt_t</code> and <code>:gt_R</code>, the <strong>ground truth</strong> visible bounding box, and the <strong>gt</strong> mask image paths.</li><li><code>test_targets</code>: the <strong>estimated</strong> bounding box, and the <strong>estimated</strong> segmentation masks.</li></ul><p>Iterate each row and load cropped images via:</p><ul><li><code>load_color_img(row, width, height)</code> </li><li><code>load_depth_img(row, width, height)</code> scaled in meters as Float32</li><li><code>load_mask_img(row, width, height)</code> masks the visible object surface either gt from disk or from the detections file in the test targets.</li></ul><h1 id="Maximum-Distance-of-Model-Points-for-indistinguishable-views-(MDD-S)"><a class="docs-heading-anchor" href="#Maximum-Distance-of-Model-Points-for-indistinguishable-views-(MDD-S)">Maximum Distance of Model Points for indistinguishable views (MDD-S)</a><a id="Maximum-Distance-of-Model-Points-for-indistinguishable-views-(MDD-S)-1"></a><a class="docs-heading-anchor-permalink" href="#Maximum-Distance-of-Model-Points-for-indistinguishable-views-(MDD-S)" title="Permalink"></a></h1><p>To avoid defining symmetries and the influence of the mesh sampling, we implement the MDD-S by replacing the <span>$avg$</span> in ADD-S with <span>$max$</span>:</p><p class="math-container">\[e_{MDDS}(\hat{\mathbf{P}},\mathbf{P};\mathcal{M})= \max_{\mathbf{x}_1 \in \mathcal{M}} \min_{\mathbf{x}_2 \in \mathcal{M}} \parallel \hat{\mathbf{P}} \mathbf{x}_1 - \mathbf{P} \mathbf{x}_2 \parallel_2\]</p><p>As (<a href="https://doi.org/10.1007/978-3-319-49409-8_52">Hodan et. al 2016</a>) describe ADD-S as the lower bound of the ADD error. However, the symmetric transformation is a subtle difference which results in MDD-S not being a lower bound of MSSD. It heavily depends on the set of global symmetries used and makes comparability impossible.</p><p><img src="assets/mdds_no_bound.jpg" alt="MDD-S not lower boud of MDDS"/></p></article><nav class="docs-footer"><a class="docs-footer-nextpage" href="methods/">Methods »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="auto">Automatic (OS)</option><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.4.0 on <span class="colophon-date" title="Wednesday 24 April 2024 14:55">Wednesday 24 April 2024</span>. Using Julia version 1.10.2.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
<html lang="en"><head><meta charset="UTF-8"/><meta name="viewport" content="width=device-width, initial-scale=1.0"/><title>PoseErrors.jl · PoseErrors</title><meta name="title" content="PoseErrors.jl · PoseErrors"/><meta property="og:title" content="PoseErrors.jl · PoseErrors"/><meta property="twitter:title" content="PoseErrors.jl · PoseErrors"/><meta name="description" content="Documentation for PoseErrors."/><meta property="og:description" content="Documentation for PoseErrors."/><meta property="twitter:description" content="Documentation for PoseErrors."/><script data-outdated-warner src="assets/warner.js"></script><link href="https://cdnjs.cloudflare.com/ajax/libs/lato-font/3.0.0/css/lato-font.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/juliamono/0.050/juliamono.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/fontawesome.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/solid.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.2/css/brands.min.css" rel="stylesheet" type="text/css"/><link href="https://cdnjs.cloudflare.com/ajax/libs/KaTeX/0.16.8/katex.min.css" rel="stylesheet" type="text/css"/><script>documenterBaseURL="."</script><script src="https://cdnjs.cloudflare.com/ajax/libs/require.js/2.3.6/require.min.js" data-main="assets/documenter.js"></script><script src="search_index.js"></script><script src="siteinfo.js"></script><script src="../versions.js"></script><link class="docs-theme-link" rel="stylesheet" type="text/css" href="assets/themes/documenter-dark.css" data-theme-name="documenter-dark" data-theme-primary-dark/><link class="docs-theme-link" rel="stylesheet" type="text/css" href="assets/themes/documenter-light.css" data-theme-name="documenter-light" data-theme-primary/><script src="assets/themeswap.js"></script></head><body><div id="documenter"><nav class="docs-sidebar"><div class="docs-package-name"><span class="docs-autofit"><a href>PoseErrors</a></span></div><button class="docs-search-query input is-rounded is-small is-clickable my-2 mx-auto py-1 px-2" id="documenter-search-query">Search docs (Ctrl + /)</button><ul class="docs-menu"><li class="is-active"><a class="tocitem" href>PoseErrors.jl</a><ul class="internal"><li class="toplevel"><a class="tocitem" href="#Setup-for-BOP-dataset-evaluation"><span>Setup for BOP dataset evaluation</span></a></li><li><a class="tocitem" href="#Loading-the-BOP-datasets"><span>Loading the BOP datasets</span></a></li><li class="toplevel"><a class="tocitem" href="#Maximum-Distance-of-Model-Points-for-indistinguishable-views-(MDD-S)"><span>Maximum Distance of Model Points for indistinguishable views (MDD-S)</span></a></li></ul></li><li><a class="tocitem" href="methods/">Methods</a></li></ul><div class="docs-version-selector field has-addons"><div class="control"><span class="docs-label button is-static is-size-7">Version</span></div><div class="docs-selector control is-expanded"><div class="select is-fullwidth is-size-7"><select id="documenter-version-selector"></select></div></div></div></nav><div class="docs-main"><header class="docs-navbar"><a class="docs-sidebar-button docs-navbar-link fa-solid fa-bars is-hidden-desktop" id="documenter-sidebar-button" href="#"></a><nav class="breadcrumb"><ul class="is-hidden-mobile"><li class="is-active"><a href>PoseErrors.jl</a></li></ul><ul class="is-hidden-tablet"><li class="is-active"><a href>PoseErrors.jl</a></li></ul></nav><div class="docs-right"><a class="docs-navbar-link" href="https://github.com/rwth-irt/PoseErrors.jl" title="View the repository on GitHub"><span class="docs-icon fa-brands"></span><span class="docs-label is-hidden-touch">GitHub</span></a><a class="docs-navbar-link" href="https://github.com/rwth-irt/PoseErrors.jl/blob/main/docs/src/index.md" title="Edit source on GitHub"><span class="docs-icon fa-solid"></span></a><a class="docs-settings-button docs-navbar-link fa-solid fa-gear" id="documenter-settings-button" href="#" title="Settings"></a><a class="docs-article-toggle-button fa-solid fa-chevron-up" id="documenter-article-toggle-button" href="javascript:;" title="Collapse all docstrings"></a></div></header><article class="content" id="documenter-page"><h1 id="PoseErrors.jl"><a class="docs-heading-anchor" href="#PoseErrors.jl">PoseErrors.jl</a><a id="PoseErrors.jl-1"></a><a class="docs-heading-anchor-permalink" href="#PoseErrors.jl" title="Permalink"></a></h1><p>A good overview and rationale behind 6D pose error metrics can be found in the <a href="https://bop.felk.cvut.cz/challenges/bop-challenge-2019/#evaluationmethodology">BOP-challenge</a>. They prefer Maximum Symmetry-Aware Surface Distance (MSSD) over Averade Average Distance of Model Points with indistinguishable views (ADD-S = ADI), because they can yield low errors with bad visual alignment. Moreover, ADD-S is dominated by higher-frequency surface parts (e.g. a Thread). Maximum distances do not suffer from the surface sampling density as much.</p><p>However, annotating the symmetries is tedious and heavily depends on the choice of coordinate frames. For a large set of objects like surgical instruments which are not exported in a standardized / symmetry aligned frame, this is impractical. So, similar to (<a href="https://doi.org/10.3390/jimaging8030053">Gorschlüter et al. 2022</a>) we use ADD-S and VDS (<a href="https://doi.org/10.1007/978-3-319-49409-8_52">Hodan et. al 2016</a>) as metrics.</p><h1 id="Setup-for-BOP-dataset-evaluation"><a class="docs-heading-anchor" href="#Setup-for-BOP-dataset-evaluation">Setup for BOP dataset evaluation</a><a id="Setup-for-BOP-dataset-evaluation-1"></a><a class="docs-heading-anchor-permalink" href="#Setup-for-BOP-dataset-evaluation" title="Permalink"></a></h1><p>Extract the BOP datasets as described on their <a href="https://bop.felk.cvut.cz/datasets/">website</a>. Move the detections JSON to the matching datasets test directory and rename it to <code>default_detections.json</code>, e.g. <em>datasets/tless/default_detections.json</em>.</p><p>You could also you the keyword argument <code>detections_file</code> of <code>scene_test_targets</code> to specify another file in the test directory.</p><h2 id="Loading-the-BOP-datasets"><a class="docs-heading-anchor" href="#Loading-the-BOP-datasets">Loading the BOP datasets</a><a id="Loading-the-BOP-datasets-1"></a><a class="docs-heading-anchor-permalink" href="#Loading-the-BOP-datasets" title="Permalink"></a></h2><p>Your first entry points should be the methods to load the according targets. These load the required image files, mesh files, camera parameters, etc.</p><ul><li><code>gt_targets</code>: Loads the ground truth pose as <code>:gt_t</code> and <code>:gt_R</code>, the <strong>ground truth</strong> visible bounding box, and the <strong>gt</strong> mask image paths.</li><li><code>test_targets</code>: the <strong>estimated</strong> bounding box, and the <strong>estimated</strong> segmentation masks.</li></ul><p>Iterate each row and load cropped images via:</p><ul><li><code>load_color_img(row, width, height)</code> </li><li><code>load_depth_img(row, width, height)</code> scaled in meters as Float32</li><li><code>load_mask_img(row, width, height)</code> masks the visible object surface either gt from disk or from the detections file in the test targets.</li></ul><h1 id="Maximum-Distance-of-Model-Points-for-indistinguishable-views-(MDD-S)"><a class="docs-heading-anchor" href="#Maximum-Distance-of-Model-Points-for-indistinguishable-views-(MDD-S)">Maximum Distance of Model Points for indistinguishable views (MDD-S)</a><a id="Maximum-Distance-of-Model-Points-for-indistinguishable-views-(MDD-S)-1"></a><a class="docs-heading-anchor-permalink" href="#Maximum-Distance-of-Model-Points-for-indistinguishable-views-(MDD-S)" title="Permalink"></a></h1><p>To avoid defining symmetries and the influence of the mesh sampling, we implement the MDD-S by replacing the <span>$avg$</span> in ADD-S with <span>$max$</span>:</p><p class="math-container">\[e_{MDDS}(\hat{\mathbf{P}},\mathbf{P};\mathcal{M})= \max_{\mathbf{x}_1 \in \mathcal{M}} \min_{\mathbf{x}_2 \in \mathcal{M}} \parallel \hat{\mathbf{P}} \mathbf{x}_1 - \mathbf{P} \mathbf{x}_2 \parallel_2\]</p><p>As (<a href="https://doi.org/10.1007/978-3-319-49409-8_52">Hodan et. al 2016</a>) describe ADD-S as the lower bound of the ADD error. However, the symmetric transformation is a subtle difference which results in MDD-S not being a lower bound of MSSD. It heavily depends on the set of global symmetries used and makes comparability impossible.</p><p><img src="assets/mdds_no_bound.jpg" alt="MDD-S not lower boud of MDDS"/></p></article><nav class="docs-footer"><a class="docs-footer-nextpage" href="methods/">Methods »</a><div class="flexbox-break"></div><p class="footer-message">Powered by <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> and the <a href="https://julialang.org/">Julia Programming Language</a>.</p></nav></div><div class="modal" id="documenter-settings"><div class="modal-background"></div><div class="modal-card"><header class="modal-card-head"><p class="modal-card-title">Settings</p><button class="delete"></button></header><section class="modal-card-body"><p><label class="label">Theme</label><div class="select"><select id="documenter-themepicker"><option value="auto">Automatic (OS)</option><option value="documenter-light">documenter-light</option><option value="documenter-dark">documenter-dark</option></select></div></p><hr/><p>This document was generated with <a href="https://github.com/JuliaDocs/Documenter.jl">Documenter.jl</a> version 1.4.0 on <span class="colophon-date" title="Wednesday 10 July 2024 07:42">Wednesday 10 July 2024</span>. Using Julia version 1.10.4.</p></section><footer class="modal-card-foot"></footer></div></div></div></body></html>
Loading

0 comments on commit 030e3b2

Please sign in to comment.