[
{
"id": "erler_2024_ppsurf",
"type_id": "journalpaper",
"tu_id": null,
"repositum_id": "20.500.12708/192690",
"title": "PPSurf: Combining Patches and Point Convolutions for Detailed Surface Reconstruction",
"date": "2024-01-12",
"abstract": "Abstract 3D surface reconstruction from point clouds is a key step in areas such as content creation, archaeology, digital cultural heritage and engineering. Current approaches either try to optimize a non-data-driven surface representation to fit the points, or learn a data-driven prior over the distribution of commonly occurring surfaces and how they correlate with potentially noisy point clouds. Data-driven methods enable robust handling of noise and typically either focus on a global or a local prior, which trade-off between robustness to noise on the global end and surface detail preservation on the local end. We propose PPSurf as a method that combines a global prior based on point convolutions and a local prior based on processing local point cloud patches. We show that this approach is robust to noise while recovering surface details more accurately than the current state-of-the-art. Our source code, pre-trained model and dataset are available at https://github.com/cg-tuwien/ppsurf.",
"authors_et_al": false,
"substitute": null,
"main_image": {
"description": "PPSurf teaser with comparison",
"filetitle": "teaser",
"main_file": true,
"use_in_gallery": true,
"access": "public",
"image_width": 2288,
"image_height": 740,
"name": "erler_2024_ppsurf-teaser.png",
"type": "image/png",
"size": 1734284,
"path": "Publication:erler_2024_ppsurf",
"url": "https://www.cg.tuwien.ac.at/research/publications/2024/erler_2024_ppsurf/erler_2024_ppsurf-teaser.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2024/erler_2024_ppsurf/erler_2024_ppsurf-teaser:thumb{{size}}.png"
},
"sync_repositum_override": "number,volume",
"repositum_presentation_id": null,
"authors": [
1395,
5317,
1919,
627,
1184,
193
],
"date_from": "2020",
"date_to": "2024-01-12",
"doi": "https://doi.org/10.1111/cgf.15000",
"event": "Eurographics 2024",
"first_published": "2024-01-12",
"html_block": "Demo
\n \n",
"issn": "1467-8659",
"journal": "Computer Graphics Forum",
"lecturer": [
1395
],
"number": "1",
"open_access": "yes",
"pages": "12",
"pages_from": "tbd",
"pages_to": "tbd",
"publisher": "WILEY",
"volume": "43",
"research_areas": [
"Geometry",
"Modeling"
],
"keywords": [
"modeling",
"surface reconstruction"
],
"weblinks": [
{
"href": "https://huggingface.co/spaces/perler/ppsurf",
"caption": "Live System",
"description": null,
"main_file": 1
},
{
"href": "https://github.com/cg-tuwien/ppsurf",
"caption": "Repo (Github)",
"description": null,
"main_file": 1
},
{
"href": "https://onlinelibrary.wiley.com/doi/10.1111/cgf.15000",
"caption": "Official Publication (Wiley Computer Graphics Forum)",
"description": null,
"main_file": 0
},
{
"href": "https://arxiv.org/abs/2401.08518",
"caption": "Preprint (ArXiv)",
"description": null,
"main_file": 0
},
{
"href": "https://www.replicabilitystamp.org/#https-github-com-cg-tuwien-ppsurf",
"caption": "Graphics Replicability Stamp Initiative",
"description": null,
"main_file": 0
}
],
"files": [
{
"description": "Paper Reproduction Code and Models\nNote: use the repo instead of this messy code",
"filetitle": "paper_repro",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler_2024_ppsurf-paper_repro.7z",
"type": "application/octet-stream",
"size": 1512332777,
"path": "Publication:erler_2024_ppsurf",
"url": "https://www.cg.tuwien.ac.at/research/publications/2024/erler_2024_ppsurf/erler_2024_ppsurf-paper_repro.7z",
"thumb_image_sizes": []
},
{
"description": "PPSurf (ArXiv Version)",
"filetitle": "paper",
"main_file": true,
"use_in_gallery": false,
"access": "public",
"name": "erler_2024_ppsurf-paper.pdf",
"type": "application/pdf",
"size": 3581909,
"path": "Publication:erler_2024_ppsurf",
"url": "https://www.cg.tuwien.ac.at/research/publications/2024/erler_2024_ppsurf/erler_2024_ppsurf-paper.pdf",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2024/erler_2024_ppsurf/erler_2024_ppsurf-paper:thumb{{size}}.png"
},
{
"description": "PPSurf 50NN Model Checkpoint",
"filetitle": "ppsurf_50nn_model",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler_2024_ppsurf-ppsurf_50nn_model.zip",
"type": "application/x-zip-compressed",
"size": 153351534,
"path": "Publication:erler_2024_ppsurf",
"url": "https://www.cg.tuwien.ac.at/research/publications/2024/erler_2024_ppsurf/erler_2024_ppsurf-ppsurf_50nn_model.zip",
"thumb_image_sizes": []
},
{
"description": "PPSurf 50NN Results (Meshes and Tables)",
"filetitle": "ppsurf_50nn_results",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler_2024_ppsurf-ppsurf_50nn_results.7z",
"type": "application/octet-stream",
"size": 1609430340,
"path": "Publication:erler_2024_ppsurf",
"url": "https://www.cg.tuwien.ac.at/research/publications/2024/erler_2024_ppsurf/erler_2024_ppsurf-ppsurf_50nn_results.7z",
"thumb_image_sizes": []
},
{
"description": "Eurographics 2024 Slides (PDF)",
"filetitle": "slides_eg24_pdf",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler_2024_ppsurf-slides_eg24_pdf.pdf",
"type": "application/pdf",
"size": 2551903,
"path": "Publication:erler_2024_ppsurf",
"url": "https://www.cg.tuwien.ac.at/research/publications/2024/erler_2024_ppsurf/erler_2024_ppsurf-slides_eg24_pdf.pdf",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2024/erler_2024_ppsurf/erler_2024_ppsurf-slides_eg24_pdf:thumb{{size}}.png"
},
{
"description": "Eurographics 2024 Slides",
"filetitle": "slides_eg24",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler_2024_ppsurf-slides_eg24.pptx",
"type": "application/vnd.openxmlformats-officedocument.presentationml.presentation",
"size": 26027659,
"path": "Publication:erler_2024_ppsurf",
"url": "https://www.cg.tuwien.ac.at/research/publications/2024/erler_2024_ppsurf/erler_2024_ppsurf-slides_eg24.pptx",
"thumb_image_sizes": []
},
{
"description": "PPSurf teaser with comparison",
"filetitle": "teaser",
"main_file": true,
"use_in_gallery": true,
"access": "public",
"image_width": 2288,
"image_height": 740,
"name": "erler_2024_ppsurf-teaser.png",
"type": "image/png",
"size": 1734284,
"path": "Publication:erler_2024_ppsurf",
"url": "https://www.cg.tuwien.ac.at/research/publications/2024/erler_2024_ppsurf/erler_2024_ppsurf-teaser.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2024/erler_2024_ppsurf/erler_2024_ppsurf-teaser:thumb{{size}}.png"
},
{
"description": "Testsets (ABC, Famous, Thingi10k)",
"filetitle": "testsets",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler_2024_ppsurf-testsets.zip",
"type": "application/x-zip-compressed",
"size": 1879212476,
"path": "Publication:erler_2024_ppsurf",
"url": "https://www.cg.tuwien.ac.at/research/publications/2024/erler_2024_ppsurf/erler_2024_ppsurf-testsets.zip",
"thumb_image_sizes": []
},
{
"description": "ABC Training Set",
"filetitle": "trainset",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler_2024_ppsurf-trainset.zip",
"type": "application/x-zip-compressed",
"size": 2513802131,
"path": "Publication:erler_2024_ppsurf",
"url": "https://www.cg.tuwien.ac.at/research/publications/2024/erler_2024_ppsurf/erler_2024_ppsurf-trainset.zip",
"thumb_image_sizes": []
}
],
"projects_workgroups": [
"rend",
"Superhumans",
"ShapeAcquisition",
"d4388"
],
"url": "https://www.cg.tuwien.ac.at/research/publications/2024/erler_2024_ppsurf/",
"__class": "Publication"
},
{
"id": "Radwan_2021_Occ",
"type_id": "journalpaper",
"tu_id": 300508,
"repositum_id": "20.500.12708/58637",
"title": "Fast occlusion-based point cloud exploration",
"date": "2021-09",
"abstract": "Large-scale unstructured point cloud scenes can be quickly visualized without prior reconstruction by utilizing levels-of-detail structures to load an appropriate subset from out-of-core storage for rendering the current view. However, as soon as we need structures within the point cloud, e.g., for interactions between objects, the construction of state-of-the-art data structures requires O(NlogN) time for N points, which is not feasible in real time for millions of points that are possibly updated in each frame. Therefore, we propose to use a surface representation structure which trades off the (here negligible) disadvantage of single-frame use for both output-dominated and near-linear construction time in practice, exploiting the inherent 2D property of sampled surfaces in 3D. This structure tightly encompasses the assumed surface of unstructured points in a set of bounding depth intervals for each cell of a discrete 2D grid. The sorted depth samples in the structure permit fast surface queries, and on top of that an occlusion graph for the scene comes almost for free. This graph enables novel real-time user operations such as revealing partially occluded objects, or scrolling through layers of occluding objects, e.g., walls in a building. As an example application we showcase a 3D scene exploration framework that enables fast, more sophisticated interactions with point clouds rendered in real time.",
"authors_et_al": false,
"substitute": null,
"main_image": {
"description": null,
"filetitle": "",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 533,
"image_height": 746,
"name": "Radwan_2021_Occ-.png",
"type": "image/png",
"size": 645321,
"path": "Publication:Radwan_2021_Occ",
"url": "https://www.cg.tuwien.ac.at/research/publications/2021/Radwan_2021_Occ/Radwan_2021_Occ-.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2021/Radwan_2021_Occ/Radwan_2021_Occ-:thumb{{size}}.png"
},
"sync_repositum_override": "date",
"repositum_presentation_id": null,
"authors": [
923,
948,
193
],
"booktitle": "The Visual Computer",
"cfp": {
"name": "CGI-call for papers.pdf",
"type": "application/pdf",
"error": "0",
"size": "1893892",
"orig_name": "CGI-call for papers.pdf",
"ext": "pdf"
},
"date_from": "2021-09-06",
"date_to": "2021-09-10",
"doi": "10.1007/s00371-021-02243-x",
"event": "Computer Graphics International 2021",
"first_published": "2021-07-28",
"issn": "1432-2315",
"journal": "The Visual Computer Journal",
"lecturer": [
923
],
"location": "online",
"open_access": "yes",
"pages": "13",
"pages_from": "2769",
"pages_to": "2781",
"publisher": "Springer",
"volume": "37",
"research_areas": [
"Modeling"
],
"keywords": [
"Software",
"Computer Graphics and Computer-Aided Design",
"Computer Vision and Pattern Recognition"
],
"weblinks": [
{
"href": "https://rdcu.be/ctQay",
"caption": null,
"description": null,
"main_file": 0
}
],
"files": [
{
"description": null,
"filetitle": "",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 533,
"image_height": 746,
"name": "Radwan_2021_Occ-.png",
"type": "image/png",
"size": 645321,
"path": "Publication:Radwan_2021_Occ",
"url": "https://www.cg.tuwien.ac.at/research/publications/2021/Radwan_2021_Occ/Radwan_2021_Occ-.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2021/Radwan_2021_Occ/Radwan_2021_Occ-:thumb{{size}}.png"
},
{
"description": null,
"filetitle": "paper",
"main_file": true,
"use_in_gallery": false,
"access": "public",
"name": "Radwan_2021_Occ-paper.pdf",
"type": "application/pdf",
"size": 2364366,
"path": "Publication:Radwan_2021_Occ",
"url": "https://www.cg.tuwien.ac.at/research/publications/2021/Radwan_2021_Occ/Radwan_2021_Occ-paper.pdf",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2021/Radwan_2021_Occ/Radwan_2021_Occ-paper:thumb{{size}}.png"
}
],
"projects_workgroups": [
"ShapeAcquisition"
],
"url": "https://www.cg.tuwien.ac.at/research/publications/2021/Radwan_2021_Occ/",
"__class": "Publication"
},
{
"id": "erler-2020-p2s",
"type_id": "inproceedings",
"tu_id": null,
"repositum_id": "20.500.12708/55555",
"title": "Points2Surf: Learning Implicit Surfaces from Point Clouds",
"date": "2020-10-28",
"abstract": "A key step in any scanning-based asset creation workflow is to convert unordered point clouds to a surface. Classical methods (e.g., Poisson reconstruction) start to degrade in the presence of noisy and partial scans. Hence, deep learning based methods have recently been proposed to produce complete surfaces, even from partial scans. However, such data-driven methods struggle to generalize to new shapes with large geometric and topological variations. We present Points2Surf, a novel patch-based learning framework that produces accurate surfaces directly from raw scans without normals.\n\nLearning a prior over a combination of detailed local patches and coarse global information improves generalization performance and reconstruction accuracy.\n\nOur extensive comparison on both synthetic and real data demonstrates a clear advantage of our method over state-of-the-art alternatives on previously unseen classes (on average, Points2Surf brings down reconstruction error by 30% over SPR and by 270%+ over deep learning based SotA methods) at the cost of longer computation times and a slight increase in small-scale topological noise in some cases. \nOur source code, pre-trained model, and dataset are available on: https://github.com/ErlerPhilipp/points2surf\n",
"authors_et_al": false,
"substitute": null,
"main_image": {
"description": "We present Points2Surf, a method to reconstruct an accurate implicit surface from a noisy point cloud. Unlike current data-driven surface reconstruction methods like DeepSDF and AtlasNet, it is patch-based, improves detail reconstruction, and unlike Screened Poisson Reconstruction (SPR), a learned prior of low-level patch shapes improves reconstruction accuracy. \nNote the quality of reconstructions, both geometric and topological, against the original surfaces. The ability of Points2Surf to generalize to new shapes makes it the first learning-based approach with significant generalization ability under both geometric and topological variations.",
"filetitle": "teaser",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 4161,
"image_height": 2179,
"name": "erler-2020-p2s-teaser.png",
"type": "image/png",
"size": 4130897,
"path": "Publication:erler-2020-p2s",
"url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-teaser.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-teaser:thumb{{size}}.png"
},
"sync_repositum_override": null,
"repositum_presentation_id": null,
"authors": [
1395,
627,
948,
193,
1771
],
"address": "Cham",
"booktitle": "Computer Vision -- ECCV 2020",
"cfp": {
"name": "eccv_2020_call_for_papers.pdf",
"type": "application/pdf",
"error": "0",
"size": "255221",
"orig_name": "eccv_2020_call_for_papers.pdf",
"ext": "pdf"
},
"date_from": "2020-08-24",
"date_to": "2020-08-27",
"doi": "10.1007/978-3-030-58558-7_7",
"editor": "Vedaldi, Andrea and Bischof, Horst and Brox, Thomas and Frahm, Jan-Michael",
"event": "ECCV 2020",
"first_published": "2020-10-28",
"isbn": "978-3-030-58558-7",
"journal": "Computer Vision – ECCV 2020",
"lecturer": [
1395
],
"location": "Glasgow, UK (online)",
"open_access": "yes",
"pages": "17",
"pages_from": "108",
"pages_to": "124",
"publisher": "Springer International Publishing",
"series": "Lecture Notes in Computer Science",
"volume": "12350",
"research_areas": [
"Geometry",
"Modeling"
],
"keywords": [
"surface reconstruction",
"implicit surfaces",
"point clouds",
"patch-based",
"local and global",
"deep learning",
"generalization"
],
"weblinks": [
{
"href": "https://github.com/ErlerPhilipp/points2surf",
"caption": "Github Repo",
"description": "Access the source code, pre-trained models and datasets from Github.",
"main_file": 0
},
{
"href": "https://arxiv.org/abs/2007.10453",
"caption": "Arxiv Pre-Print",
"description": "Access the paper pre-print on arXiv.",
"main_file": 0
},
{
"href": "https://link.springer.com/chapter/10.1007%2F978-3-030-58558-7_7",
"caption": "Official Publication",
"description": "Access the paper Springer Link.",
"main_file": 0
}
],
"files": [
{
"description": "Our normalized ground-truth meshes from the ABC dataset.",
"filetitle": "abc_meshes",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler-2020-p2s-abc_meshes.zip",
"type": "application/x-zip-compressed",
"size": 726933282,
"path": "Publication:erler-2020-p2s",
"url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-abc_meshes.zip",
"thumb_image_sizes": []
},
{
"description": "Our processed extract from the ABC dataset that can be used to train Points2Surf.",
"filetitle": "abc_training",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler-2020-p2s-abc_training.zip",
"type": "application/x-zip-compressed",
"size": 1786245314,
"path": "Publication:erler-2020-p2s",
"url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-abc_training.zip",
"thumb_image_sizes": []
},
{
"description": "Our processed extract from the ABC dataset that can be used to replicate the results of Points2Surf.",
"filetitle": "abc",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler-2020-p2s-abc.zip",
"type": "application/x-zip-compressed",
"size": 138638856,
"path": "Publication:erler-2020-p2s",
"url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-abc.zip",
"thumb_image_sizes": []
},
{
"description": "Our pre-trained models (ablation versions) to reproduce the results.",
"filetitle": "ablation_models",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler-2020-p2s-ablation_models.zip",
"type": "application/x-zip-compressed",
"size": 242492525,
"path": "Publication:erler-2020-p2s",
"url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-ablation_models.zip",
"thumb_image_sizes": []
},
{
"description": "Our processed Famous dataset that can be used to replicate the results of Points2Surf.",
"filetitle": "famous",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler-2020-p2s-famous.zip",
"type": "application/x-zip-compressed",
"size": 1525449349,
"path": "Publication:erler-2020-p2s",
"url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-famous.zip",
"thumb_image_sizes": []
},
{
"description": "8 min video for ECCV 2020",
"filetitle": "long video",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler-2020-p2s-long video.mp4",
"type": "video/mp4",
"size": 85439106,
"path": "Publication:erler-2020-p2s",
"url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-long video.mp4",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-long video:thumb{{size}}.png",
"video_mp4": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-long video:video.mp4"
},
{
"description": "A new model version with even better results.",
"filetitle": "max_model",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler-2020-p2s-max_model.zip",
"type": "application/x-zip-compressed",
"size": 20010604,
"path": "Publication:erler-2020-p2s",
"url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-max_model.zip",
"thumb_image_sizes": []
},
{
"description": "Points2Surf paper (Arxiv version)",
"filetitle": "points2surf_paper",
"main_file": true,
"use_in_gallery": false,
"access": "public",
"name": "erler-2020-p2s-points2surf_paper.pdf",
"type": "application/pdf",
"size": 9135835,
"path": "Publication:erler-2020-p2s",
"url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-points2surf_paper.pdf",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-points2surf_paper:thumb{{size}}.png"
},
{
"description": "Our processed real-world dataset that can be used to replicate the results of Points2Surf.",
"filetitle": "real_world",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler-2020-p2s-real_world.zip",
"type": "application/x-zip-compressed",
"size": 5942947,
"path": "Publication:erler-2020-p2s",
"url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-real_world.zip",
"thumb_image_sizes": []
},
{
"description": "2 min video for ECCV 2020",
"filetitle": "short video",
"main_file": true,
"use_in_gallery": false,
"access": "public",
"name": "erler-2020-p2s-short video.mp4",
"type": "video/mp4",
"size": 13862480,
"path": "Publication:erler-2020-p2s",
"url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-short video.mp4",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-short video:thumb{{size}}.png",
"video_mp4": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-short video:video.mp4"
},
{
"description": "We present Points2Surf, a method to reconstruct an accurate implicit surface from a noisy point cloud. Unlike current data-driven surface reconstruction methods like DeepSDF and AtlasNet, it is patch-based, improves detail reconstruction, and unlike Screened Poisson Reconstruction (SPR), a learned prior of low-level patch shapes improves reconstruction accuracy. \nNote the quality of reconstructions, both geometric and topological, against the original surfaces. The ability of Points2Surf to generalize to new shapes makes it the first learning-based approach with significant generalization ability under both geometric and topological variations.",
"filetitle": "teaser",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 4161,
"image_height": 2179,
"name": "erler-2020-p2s-teaser.png",
"type": "image/png",
"size": 4130897,
"path": "Publication:erler-2020-p2s",
"url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-teaser.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-teaser:thumb{{size}}.png"
},
{
"description": "Our processed extract from the Thingi10k dataset that can be used to replicate the results of Points2Surf.",
"filetitle": "thingi10k",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler-2020-p2s-thingi10k.zip",
"type": "application/x-zip-compressed",
"size": 940826980,
"path": "Publication:erler-2020-p2s",
"url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-thingi10k.zip",
"thumb_image_sizes": []
},
{
"description": "Our pre-trained model (vanilla version) to reproduce the results.",
"filetitle": "vanilla_model",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"name": "erler-2020-p2s-vanilla_model.zip",
"type": "application/x-zip-compressed",
"size": 23009153,
"path": "Publication:erler-2020-p2s",
"url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/erler-2020-p2s-vanilla_model.zip",
"thumb_image_sizes": []
}
],
"projects_workgroups": [
"Superhumans",
"MAKE-IT-FAB",
"ShapeAcquisition"
],
"url": "https://www.cg.tuwien.ac.at/research/publications/2020/erler-2020-p2s/",
"__class": "Publication"
},
{
"id": "gruber_horst-2019-baa",
"type_id": "bachelorthesis",
"tu_id": null,
"repositum_id": null,
"title": "Extensible Image Classification",
"date": "2019-08-16",
"abstract": "Struktur des Systems:\nInput: Bereits segmentierte Objekte\n1) Labeling: Input: Objekt(e), Output: Manuelle Zuordnung eines (neuen bzw. korrigierten) Labels pro Objekt\n2) Clustering: Input: Cluster mit Objekten, mit k (mindestens, oder genau zwei) verschiedenen Label, Unterteilung durch k-means Clustering, Output: k Cluster von 'ähnlichen' Objekten pro Label\n3) Training: Input: Objekte mit Label (ihres Clusters), Output: Network, das die Objekte in die derzeit gelabelten Cluster klassifiziert\nDann geht's iterativ mit Schritt 1 weiter - Objekte werden dem User angezeigt und bei Bedarf relabeled, wonach Schritt 2 folgt usw.",
"authors_et_al": false,
"substitute": null,
"main_image": {
"description": null,
"filetitle": "image",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 1008,
"image_height": 502,
"name": "gruber_horst-2019-baa-image.jpg",
"type": "image/jpeg",
"size": 189279,
"path": "Publication:gruber_horst-2019-baa",
"url": "https://www.cg.tuwien.ac.at/research/publications/2019/gruber_horst-2019-baa/gruber_horst-2019-baa-image.jpg",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2019/gruber_horst-2019-baa/gruber_horst-2019-baa-image:thumb{{size}}.png"
},
"sync_repositum_override": null,
"repositum_presentation_id": null,
"authors": [
1645
],
"co_supervisor": [
948
],
"date_end": "2019-08-16",
"date_start": "2019-04-11",
"matrikelnr": "8330864",
"supervisor": [
948
],
"research_areas": [
"Geometry"
],
"keywords": [
"deep learning",
"ontology",
"image classification"
],
"weblinks": [],
"files": [
{
"description": null,
"filetitle": "image",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 1008,
"image_height": 502,
"name": "gruber_horst-2019-baa-image.jpg",
"type": "image/jpeg",
"size": 189279,
"path": "Publication:gruber_horst-2019-baa",
"url": "https://www.cg.tuwien.ac.at/research/publications/2019/gruber_horst-2019-baa/gruber_horst-2019-baa-image.jpg",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2019/gruber_horst-2019-baa/gruber_horst-2019-baa-image:thumb{{size}}.png"
}
],
"projects_workgroups": [
"ShapeAcquisition"
],
"url": "https://www.cg.tuwien.ac.at/research/publications/2019/gruber_horst-2019-baa/",
"__class": "Publication"
},
{
"id": "ohrhallinger_stefan-2018-cgf",
"type_id": "journalpaper",
"tu_id": null,
"repositum_id": null,
"title": "FitConnect: Connecting Noisy 2D Samples by Fitted Neighborhoods",
"date": "2019-02",
"abstract": "We propose a parameter-free method to recover manifold connectivity in unstructured 2D point clouds with high noise in terms of the local feature size. This enables us to capture the features which emerge out of the noise. To achieve this, we extend the reconstruction algorithm HNN-Crust, which connects samples to two (noise-free) neighbors and has been proven to output a manifold for a relaxed sampling condition. Applying this condition to noisy samples by projecting their k-nearest neighborhoods onto local circular fits leads to multiple candidate neighbor pairs and thus makes connecting them consistently an NP-hard problem. To solve this efficiently, we design an algorithm that searches that solution space iteratively on different scales of k. It achieves linear time complexity in terms of point count plus quadratic time in the size of noise clusters. Our algorithm FitConnect extends HNN-Crust seamlessly to connect both samples with and without noise, performs as local as the recovered features and can output multiple open or closed piece-wise curves. Incidentally, our method simplifies the output geometry by eliminating all but a representative point from noisy clusters. Since local neighborhood fits overlap consistently, the resulting connectivity represents an ordering of the samples along a manifold. This permits us to simply blend the local fits for denoising with the locally estimated noise extent. Aside from applications like reconstructing silhouettes of noisy sensed data, this lays important groundwork to improve surface reconstruction in 3D. Our open-source algorithm is available online.",
"authors_et_al": false,
"substitute": null,
"main_image": {
"description": "Manifold curve fitted samples with highly varying noise",
"filetitle": "image",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 704,
"image_height": 704,
"name": "ohrhallinger_stefan-2018-cgf-image.png",
"type": "image/png",
"size": 12741,
"path": "Publication:ohrhallinger_stefan-2018-cgf",
"url": "https://www.cg.tuwien.ac.at/research/publications/2019/ohrhallinger_stefan-2018-cgf/ohrhallinger_stefan-2018-cgf-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2019/ohrhallinger_stefan-2018-cgf/ohrhallinger_stefan-2018-cgf-image:thumb{{size}}.png"
},
"sync_repositum_override": null,
"repositum_presentation_id": null,
"authors": [
948,
193
],
"cfp": {
"name": "SGP_2018_Poster_v5_small.pdf",
"type": "application/pdf",
"error": "0",
"size": "748681",
"orig_name": "SGP_2018_Poster_v5_small.pdf",
"ext": "pdf"
},
"date_from": "2018-07-07",
"date_to": "2018-07-11",
"doi": "10.1111/cgf.13395",
"event": "Eurographics Symposium on Geometry Processing",
"first_published": "2018-05-11",
"issn": "1467-8659",
"journal": "Computer Graphics Forum",
"lecturer": [
948
],
"location": "Paris, France",
"number": "1",
"pages_from": "126",
"pages_to": "137",
"volume": "38",
"research_areas": [
"Geometry"
],
"keywords": [
"curve fitting, noisy samples, guarantees, curve reconstruction"
],
"weblinks": [
{
"href": "https://github.com/stefango74/fitconnect",
"caption": "Replicability Source Code",
"description": null,
"main_file": 1
}
],
"files": [
{
"description": "Manifold curve fitted samples with highly varying noise",
"filetitle": "image",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 704,
"image_height": 704,
"name": "ohrhallinger_stefan-2018-cgf-image.png",
"type": "image/png",
"size": 12741,
"path": "Publication:ohrhallinger_stefan-2018-cgf",
"url": "https://www.cg.tuwien.ac.at/research/publications/2019/ohrhallinger_stefan-2018-cgf/ohrhallinger_stefan-2018-cgf-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2019/ohrhallinger_stefan-2018-cgf/ohrhallinger_stefan-2018-cgf-image:thumb{{size}}.png"
},
{
"description": "preprint",
"filetitle": "paper",
"main_file": true,
"use_in_gallery": false,
"access": "public",
"name": "ohrhallinger_stefan-2018-cgf-paper.pdf",
"type": "application/pdf",
"size": 842370,
"path": "Publication:ohrhallinger_stefan-2018-cgf",
"url": "https://www.cg.tuwien.ac.at/research/publications/2019/ohrhallinger_stefan-2018-cgf/ohrhallinger_stefan-2018-cgf-paper.pdf",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2019/ohrhallinger_stefan-2018-cgf/ohrhallinger_stefan-2018-cgf-paper:thumb{{size}}.png"
}
],
"projects_workgroups": [
"ShapeAcquisition"
],
"url": "https://www.cg.tuwien.ac.at/research/publications/2019/ohrhallinger_stefan-2018-cgf/",
"__class": "Publication"
},
{
"id": "ohrhallinger_stefan-2018-pg",
"type_id": "inproceedings",
"tu_id": null,
"repositum_id": null,
"title": "StretchDenoise: Parametric Curve Reconstruction with Guarantees by Separating Connectivity from Residual Uncertainty of Samples",
"date": "2018-08-24",
"abstract": "We reconstruct a closed denoised curve from an unstructured and highly noisy 2D point cloud.\nOur proposed method uses a two-pass approach: Previously recovered manifold connectivity is used for ordering noisy samples along this manifold and express these as residuals in order to enable parametric denoising.\nThis separates recovering low-frequency features from denoising high frequencies, which avoids over-smoothing.\nThe noise probability density functions (PDFs) at samples are either taken from sensor noise models or from estimates of the connectivity recovered in the first pass.\nThe output curve balances the signed distances (inside/outside) to the samples.\nAdditionally, the angles between edges of the polygon representing the connectivity become minimized in the least-square sense.\nThe movement of the polygon's vertices is restricted to their noise extent, i.e., a cut-off distance corresponding to a maximum variance of the PDFs.\nWe approximate the resulting optimization model, which consists of higher-order functions, by a linear model with good correspondence.\nOur algorithm is parameter-free and operates fast on the local neighborhoods determined by the connectivity.\n%We augment a least-squares solver constrained by a linear system to also handle bounds.\nThis enables us to guarantee stochastic error bounds for sampled curves corrupted by noise, e.g., silhouettes from sensed data, and we improve on the reconstruction error from ground truth.\nSource code is available online. An extended version is available at: https://arxiv.org/abs/1808.07778",
"authors_et_al": false,
"substitute": null,
"main_image": {
"description": null,
"filetitle": "image",
"main_file": true,
"use_in_gallery": true,
"access": "public",
"image_width": 828,
"image_height": 812,
"name": "ohrhallinger_stefan-2018-pg-image.png",
"type": "image/png",
"size": 68562,
"path": "Publication:ohrhallinger_stefan-2018-pg",
"url": "https://www.cg.tuwien.ac.at/research/publications/2018/ohrhallinger_stefan-2018-pg/ohrhallinger_stefan-2018-pg-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2018/ohrhallinger_stefan-2018-pg/ohrhallinger_stefan-2018-pg-image:thumb{{size}}.png"
},
"sync_repositum_override": null,
"repositum_presentation_id": null,
"authors": [
948,
193
],
"booktitle": "Proceedings of Pacific Graphics 2018",
"date_from": "2018-10-08",
"date_to": "2018-10-11",
"editor": "H. Fu, A. Ghosh, and J. Kopf (Guest Editors)",
"event": "Pacific Graphics 2018",
"isbn": "978-3-03868-073-4",
"lecturer": [
948
],
"location": "Hong Kong",
"pages_from": "1",
"pages_to": "4",
"research_areas": [
"Geometry"
],
"keywords": [
"Denoising",
"Curve reconstruction",
"Optimization"
],
"weblinks": [
{
"href": "https://arxiv.org/abs/1808.07778",
"caption": "Extended version",
"description": null,
"main_file": 1
},
{
"href": "https://github.com/stefango74/stretchdenoise",
"caption": "source",
"description": "source code on github",
"main_file": 1
}
],
"files": [
{
"description": null,
"filetitle": "image",
"main_file": true,
"use_in_gallery": true,
"access": "public",
"image_width": 828,
"image_height": 812,
"name": "ohrhallinger_stefan-2018-pg-image.png",
"type": "image/png",
"size": 68562,
"path": "Publication:ohrhallinger_stefan-2018-pg",
"url": "https://www.cg.tuwien.ac.at/research/publications/2018/ohrhallinger_stefan-2018-pg/ohrhallinger_stefan-2018-pg-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2018/ohrhallinger_stefan-2018-pg/ohrhallinger_stefan-2018-pg-image:thumb{{size}}.png"
},
{
"description": null,
"filetitle": "paper",
"main_file": true,
"use_in_gallery": false,
"access": "public",
"name": "ohrhallinger_stefan-2018-pg-paper.pdf",
"type": "application/pdf",
"size": 703514,
"path": "Publication:ohrhallinger_stefan-2018-pg",
"url": "https://www.cg.tuwien.ac.at/research/publications/2018/ohrhallinger_stefan-2018-pg/ohrhallinger_stefan-2018-pg-paper.pdf",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2018/ohrhallinger_stefan-2018-pg/ohrhallinger_stefan-2018-pg-paper:thumb{{size}}.png"
}
],
"projects_workgroups": [
"ShapeAcquisition"
],
"url": "https://www.cg.tuwien.ac.at/research/publications/2018/ohrhallinger_stefan-2018-pg/",
"__class": "Publication"
},
{
"id": "pointner_michael-2017-baa",
"type_id": "bachelorthesis",
"tu_id": null,
"repositum_id": null,
"title": "Multi-Focal Image Generation using Automatic Depth-Based Focus Selection",
"date": "2017-09-28",
"abstract": "One of the most important objectives in photography is the sharpness of the whole\nimage, which is not so easy to achieve because of the physical properties of a camera lens.\nDue to Google’s new Tango API, which enables the perception of depth information on\nsmartphones, we have evaluated the usability of this new technology for the generation of\nall-sharp images through multi-focus with the Lenovo Phab 2 Pro as the first smartphone\nto support this technology.",
"authors_et_al": false,
"substitute": null,
"main_image": {
"description": "Image vs. depth image",
"filetitle": "image",
"main_file": true,
"use_in_gallery": true,
"access": "public",
"image_width": 510,
"image_height": 440,
"name": "pointner_michael-2017-baa-image.png",
"type": "image/png",
"size": 331005,
"path": "Publication:pointner_michael-2017-baa",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/pointner_michael-2017-baa/pointner_michael-2017-baa-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/pointner_michael-2017-baa/pointner_michael-2017-baa-image:thumb{{size}}.png"
},
"sync_repositum_override": null,
"repositum_presentation_id": null,
"authors": [
1535
],
"date_end": "2017-08-31",
"date_start": "2017-03-01",
"matrikelnr": "01427791",
"supervisor": [
948
],
"research_areas": [
"Geometry",
"Rendering"
],
"keywords": [
"all-in-focus",
"focus selection",
"image stitching"
],
"weblinks": [],
"files": [
{
"description": "Image vs. depth image",
"filetitle": "image",
"main_file": true,
"use_in_gallery": true,
"access": "public",
"image_width": 510,
"image_height": 440,
"name": "pointner_michael-2017-baa-image.png",
"type": "image/png",
"size": 331005,
"path": "Publication:pointner_michael-2017-baa",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/pointner_michael-2017-baa/pointner_michael-2017-baa-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/pointner_michael-2017-baa/pointner_michael-2017-baa-image:thumb{{size}}.png"
},
{
"description": "Bachelor thesis",
"filetitle": "thesis",
"main_file": true,
"use_in_gallery": false,
"access": "public",
"name": "pointner_michael-2017-baa-thesis.pdf",
"type": "application/pdf",
"size": 30588625,
"path": "Publication:pointner_michael-2017-baa",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/pointner_michael-2017-baa/pointner_michael-2017-baa-thesis.pdf",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/pointner_michael-2017-baa/pointner_michael-2017-baa-thesis:thumb{{size}}.png"
}
],
"projects_workgroups": [
"ShapeAcquisition"
],
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/pointner_michael-2017-baa/",
"__class": "Publication"
},
{
"id": "erler_philipp-2017-phd",
"type_id": "phdthesis",
"tu_id": null,
"repositum_id": null,
"title": "Smart Surface Reconstruction",
"date": "2017-09",
"abstract": null,
"authors_et_al": false,
"substitute": null,
"main_image": {
"description": "Reconstruction, 3D representations",
"filetitle": "image",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"image_width": 1440,
"image_height": 501,
"name": "erler_philipp-2017-phd-image.png",
"type": "image/png",
"size": 221151,
"path": "Publication:erler_philipp-2017-phd",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/erler_philipp-2017-phd/erler_philipp-2017-phd-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/erler_philipp-2017-phd/erler_philipp-2017-phd-image:thumb{{size}}.png"
},
"sync_repositum_override": null,
"repositum_presentation_id": null,
"authors": [
1395
],
"date_end": "2026-02-11",
"date_start": "2017-09",
"duration": "3 years",
"reviewer_1": [
5558
],
"reviewer_2": [
1519
],
"rigorosum": "2026-02-11",
"supervisor": [
193
],
"research_areas": [
"Geometry",
"Modeling"
],
"keywords": [
"Surface Reconstruction",
"Deep Learning",
"Point Cloud"
],
"weblinks": [],
"files": [
{
"description": "Dissertation PDF",
"filetitle": "dissertation",
"main_file": true,
"use_in_gallery": false,
"access": "public",
"name": "erler_philipp-2017-phd-dissertation.pdf",
"type": "application/pdf",
"size": 115758696,
"path": "Publication:erler_philipp-2017-phd",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/erler_philipp-2017-phd/erler_philipp-2017-phd-dissertation.pdf",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/erler_philipp-2017-phd/erler_philipp-2017-phd-dissertation:thumb{{size}}.png"
},
{
"description": "Reconstruction, 3D representations",
"filetitle": "image",
"main_file": false,
"use_in_gallery": false,
"access": "public",
"image_width": 1440,
"image_height": 501,
"name": "erler_philipp-2017-phd-image.png",
"type": "image/png",
"size": 221151,
"path": "Publication:erler_philipp-2017-phd",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/erler_philipp-2017-phd/erler_philipp-2017-phd-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/erler_philipp-2017-phd/erler_philipp-2017-phd-image:thumb{{size}}.png"
}
],
"projects_workgroups": [
"ShapeAcquisition",
"d4388",
"Superhumans"
],
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/erler_philipp-2017-phd/",
"__class": "Publication"
},
{
"id": "grossmann-2016-baa",
"type_id": "bachelorthesis",
"tu_id": null,
"repositum_id": null,
"title": "Extracting Sensor Specific Noise Models",
"date": "2017-08-28",
"abstract": "With the growing number of consumer-oriented depth sensors like the Kinect or the\nnewly released Phab2Pro, the question of how precise these sensors are arises. In this\nthesis we want to evaluate the average noise in the generated depth measurements in\nboth the axial direction and the lateral directions. As part of a two-part project this\nthesis will view the noise’s development with varying distance and angle. Finally, we\nwill present and evaluate two models describing the noise behavior, with the first being\nderived from solely this thesis’ measurements and the second one being a combination of\nthe previous model and a model of a colleague. This derived models can be used in a\npost-processing step to filter the generated depth images.",
"authors_et_al": false,
"substitute": null,
"main_image": {
"description": "Kinect v2 and Phab2Pro noise model evaluation",
"filetitle": "image",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 504,
"image_height": 606,
"name": "grossmann-2016-baa-image.png",
"type": "image/png",
"size": 217008,
"path": "Publication:grossmann-2016-baa",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/grossmann-2016-baa/grossmann-2016-baa-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/grossmann-2016-baa/grossmann-2016-baa-image:thumb{{size}}.png"
},
"sync_repositum_override": null,
"repositum_presentation_id": null,
"authors": [
1366
],
"co_supervisor": [
193
],
"date_end": "2017-08-28",
"date_start": "2016-06-17",
"duration": "4 months",
"matrikelnr": "01325103",
"note": "1",
"supervisor": [
948
],
"research_areas": [
"Geometry"
],
"keywords": [
"noise model",
"surface reconstruction",
"sensor noise"
],
"weblinks": [],
"files": [
{
"description": "Kinect v2 and Phab2Pro noise model evaluation",
"filetitle": "image",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 504,
"image_height": 606,
"name": "grossmann-2016-baa-image.png",
"type": "image/png",
"size": 217008,
"path": "Publication:grossmann-2016-baa",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/grossmann-2016-baa/grossmann-2016-baa-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/grossmann-2016-baa/grossmann-2016-baa-image:thumb{{size}}.png"
},
{
"description": "Bachelor thesis",
"filetitle": "thesis",
"main_file": true,
"use_in_gallery": false,
"access": "public",
"name": "grossmann-2016-baa-thesis.pdf",
"type": "application/pdf",
"size": 12775551,
"path": "Publication:grossmann-2016-baa",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/grossmann-2016-baa/grossmann-2016-baa-thesis.pdf",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/grossmann-2016-baa/grossmann-2016-baa-thesis:thumb{{size}}.png"
}
],
"projects_workgroups": [
"ShapeAcquisition"
],
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/grossmann-2016-baa/",
"__class": "Publication"
},
{
"id": "koeppel-2016-baa",
"type_id": "bachelorthesis",
"tu_id": null,
"repositum_id": null,
"title": "Extracting Noise Models – considering X / Y and Z Noise",
"date": "2017-08-28",
"abstract": "We have developed two different test setups allowing the characterization of noise in X,\nY and Z direction for the KinectV2 and the Phab2Pro depth sensors. We have combined\nthese two methods, generating a single noise model allowing a prediction of the amount of\nnoise in specific areas of an image in the three respective directions at a certain distance\nand rotation. We have conducted two test setups and measured the noise from 900 mm\nto 3.100 mm for the generation of the noise models. The test setup of this thesis focused\non determining the noise in X, Y and Z direction, covering the whole frustum of the\nrespective depth sensor. In this thesis, Z noise was measured against a wall and X and\nY noises were measured using a 3D chequerboard that was shifted through the room,\nallowing the above mentioned coverage of the whole frustum. Along the edges of the\ncells of the chequerboard, the X and Y noise was measured. The combined model was\nevaluated by using a solid cube to classify the quality of our noise model.",
"authors_et_al": false,
"substitute": null,
"main_image": {
"description": "Camera calibration prior to extraction of depth values",
"filetitle": "image",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 443,
"image_height": 316,
"name": "koeppel-2016-baa-image.png",
"type": "image/png",
"size": 133989,
"path": "Publication:koeppel-2016-baa",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/koeppel-2016-baa/koeppel-2016-baa-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/koeppel-2016-baa/koeppel-2016-baa-image:thumb{{size}}.png"
},
"sync_repositum_override": null,
"repositum_presentation_id": null,
"authors": [
1367
],
"co_supervisor": [
193
],
"date_end": "2017-08-28",
"date_start": "2016-06-17",
"duration": "4 months",
"matrikelnr": "01327052",
"note": "1",
"supervisor": [
948
],
"research_areas": [
"Geometry"
],
"keywords": [
"noise model",
"surface reconstruction",
"sensor noise"
],
"weblinks": [],
"files": [
{
"description": "Camera calibration prior to extraction of depth values",
"filetitle": "image",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 443,
"image_height": 316,
"name": "koeppel-2016-baa-image.png",
"type": "image/png",
"size": 133989,
"path": "Publication:koeppel-2016-baa",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/koeppel-2016-baa/koeppel-2016-baa-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/koeppel-2016-baa/koeppel-2016-baa-image:thumb{{size}}.png"
},
{
"description": "Bachelor thesis",
"filetitle": "thesis",
"main_file": true,
"use_in_gallery": false,
"access": "public",
"name": "koeppel-2016-baa-thesis.pdf",
"type": "application/pdf",
"size": 8955320,
"path": "Publication:koeppel-2016-baa",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/koeppel-2016-baa/koeppel-2016-baa-thesis.pdf",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/koeppel-2016-baa/koeppel-2016-baa-thesis:thumb{{size}}.png"
}
],
"projects_workgroups": [
"ShapeAcquisition"
],
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/koeppel-2016-baa/",
"__class": "Publication"
},
{
"id": "Radwan-2017-Occ",
"type_id": "inproceedings",
"tu_id": null,
"repositum_id": null,
"title": "Cut and Paint: Occlusion-Aware Subset Selection for Surface Processing",
"date": "2017-05",
"abstract": "User-guided surface selection operations are straightforward for visible regions on a convex model. However, concave surfaces present a challenge because self-occlusions require multiple camera positions to get unobstructed views. Therefore, users often have to locate and switch to new unobstructed views in order to continue the operation. Our novel approach enables operations like painting or cutting in a single view, even on the backside of objects and for arbitrary depth complexity, with interactive performance. Continuous projection of a curve drawn in screen space onto the mesh guarantees seamless brush strokes or manifold cuts, unaffected by any occlusions.\n\nOur occlusion-aware surface-processing method enables a number of applications in an easy way. As examples, we show continuous painting on the surface, selecting regions for texturing, creating illustrative cutaways from nested models and animation of cutaways.",
"authors_et_al": false,
"substitute": null,
"main_image": {
"description": "Occlusion-aware generation of a cutaway illustration.",
"filetitle": "image",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 1964,
"image_height": 2040,
"name": "Radwan-2017-Occ-image.jpg",
"type": "image/jpeg",
"size": 735971,
"path": "Publication:Radwan-2017-Occ",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/Radwan-2017-Occ/Radwan-2017-Occ-image.jpg",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/Radwan-2017-Occ/Radwan-2017-Occ-image:thumb{{size}}.png"
},
"sync_repositum_override": null,
"repositum_presentation_id": null,
"authors": [
923,
948,
823,
193
],
"booktitle": "Proceedings of Graphics Interface 2017",
"date_from": "2017-05-17",
"date_to": "2017-05-19",
"doi": "10.20380/GI2017.11",
"event": "Graphics Interface 2017",
"lecturer": [
923
],
"location": "Edmonton, Alberta, CA",
"pages_from": "82",
"pages_to": "89",
"publisher": "Canadian Human-Computer Communications Society / Soci{\\'e}t{\\'e} canadienne du dialogue humain-machine",
"research_areas": [
"Geometry",
"IllVis"
],
"keywords": [],
"weblinks": [],
"files": [
{
"description": "Occlusion-aware generation of a cutaway illustration.",
"filetitle": "image",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 1964,
"image_height": 2040,
"name": "Radwan-2017-Occ-image.jpg",
"type": "image/jpeg",
"size": 735971,
"path": "Publication:Radwan-2017-Occ",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/Radwan-2017-Occ/Radwan-2017-Occ-image.jpg",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/Radwan-2017-Occ/Radwan-2017-Occ-image:thumb{{size}}.png"
},
{
"description": null,
"filetitle": "paper",
"main_file": true,
"use_in_gallery": false,
"access": "public",
"name": "Radwan-2017-Occ-paper.pdf",
"type": "application/pdf",
"size": 5813944,
"path": "Publication:Radwan-2017-Occ",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/Radwan-2017-Occ/Radwan-2017-Occ-paper.pdf",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/Radwan-2017-Occ/Radwan-2017-Occ-paper:thumb{{size}}.png"
}
],
"projects_workgroups": [
"ShapeAcquisition"
],
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/Radwan-2017-Occ/",
"__class": "Publication"
},
{
"id": "reinwald-2017-baa",
"type_id": "bachelorthesis",
"tu_id": null,
"repositum_id": null,
"title": "Fast kNN in Screen Space on GPU",
"date": "2017-03-20",
"abstract": "A common problem in computer science is to construct a spatial data structure (octree, kd-tree) and use it to search for k-nearest neighbors (kNN). In surface reconstruction from dynamic points (real-time), both construction and search time are critical. As points on a surface are a sparse distribution in 3D, this can be exploited by mapping them into screen space (2D), as shown in \"Auto-splats\" (Preiner 2012). Our approach is to also exploit spatial coherence in screen space to find kNN for points. The performance is maximized by a CUDA implementation designed to minimize memory-boundness. An preliminary implementation exists which constructs a packed quadtree and reads kNN from a quadtree node density estimate. A few improvements have to be made to optimize performance and to analyze results.",
"authors_et_al": false,
"substitute": null,
"main_image": {
"description": null,
"filetitle": "image",
"main_file": true,
"use_in_gallery": true,
"access": "public",
"image_width": 637,
"image_height": 602,
"name": "reinwald-2017-baa-image.png",
"type": "image/png",
"size": 398232,
"path": "Publication:reinwald-2017-baa",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/reinwald-2017-baa/reinwald-2017-baa-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/reinwald-2017-baa/reinwald-2017-baa-image:thumb{{size}}.png"
},
"sync_repositum_override": null,
"repositum_presentation_id": null,
"authors": [
1420
],
"date_end": "2019-03-12",
"date_start": "2017-03-21",
"matrikelnr": "1126981",
"note": "1",
"supervisor": [
948
],
"research_areas": [
"Geometry"
],
"keywords": [
"k-nearest neighbors",
"surface reconstruction",
"parallel optimization"
],
"weblinks": [],
"files": [
{
"description": null,
"filetitle": "image",
"main_file": true,
"use_in_gallery": true,
"access": "public",
"image_width": 637,
"image_height": 602,
"name": "reinwald-2017-baa-image.png",
"type": "image/png",
"size": 398232,
"path": "Publication:reinwald-2017-baa",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/reinwald-2017-baa/reinwald-2017-baa-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/reinwald-2017-baa/reinwald-2017-baa-image:thumb{{size}}.png"
},
{
"description": null,
"filetitle": "thesis",
"main_file": true,
"use_in_gallery": false,
"access": "public",
"name": "reinwald-2017-baa-thesis.pdf",
"type": "application/pdf",
"size": 2988235,
"path": "Publication:reinwald-2017-baa",
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/reinwald-2017-baa/reinwald-2017-baa-thesis.pdf",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2017/reinwald-2017-baa/reinwald-2017-baa-thesis:thumb{{size}}.png"
}
],
"projects_workgroups": [
"ShapeAcquisition"
],
"url": "https://www.cg.tuwien.ac.at/research/publications/2017/reinwald-2017-baa/",
"__class": "Publication"
},
{
"id": "forsythe-2016-ccm",
"type_id": "journalpaper",
"tu_id": null,
"repositum_id": null,
"title": "Resolution-independent superpixels based on convex constrained meshes without small angles",
"date": "2016-12",
"abstract": "The over-segmentation problem for images is studied in the new resolution-independent formulation when a large image is approximated by a small number of convex polygons with straight edges at subpixel precision. These polygonal superpixels are obtained by refining and extending subpixel edge segments to a full mesh of convex polygons without small angles and with approximation guarantees. Another novelty is the objective error difference between an original pixel-based image and the reconstructed image with a best constant color over each superpixel, which does not need human segmentations. The experiments on images from the Berkeley Segmentation Database show that new meshes are smaller and provide better approximations than the state-of-the-art.",
"authors_et_al": false,
"substitute": null,
"main_image": {
"description": "null",
"filetitle": "image",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 512,
"image_height": 512,
"name": "forsythe-2016-ccm-image.png",
"type": "image/png",
"size": 401897,
"path": "Publication:forsythe-2016-ccm",
"url": "https://www.cg.tuwien.ac.at/research/publications/2016/forsythe-2016-ccm/forsythe-2016-ccm-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2016/forsythe-2016-ccm/forsythe-2016-ccm-image:thumb{{size}}.png"
},
"sync_repositum_override": null,
"repositum_presentation_id": null,
"authors": [
1402,
1403,
1404
],
"date_from": "2016-12-12",
"date_to": "2016-12-14",
"event": "ISVC 2016",
"issn": "0302-9743",
"journal": "Lecture Notes in Computer Science (LNCS)",
"lecturer": [
1402
],
"location": "Las Vegas",
"pages_from": "223",
"pages_to": "233",
"protocol": "null",
"volume": "10072",
"research_areas": [
"Geometry"
],
"keywords": [
"superpixels",
"polygonal mesh",
"Delaunay triangulation",
"constrained triangulation",
"edge detection"
],
"weblinks": [],
"files": [
{
"description": "null",
"filetitle": "image",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 512,
"image_height": 512,
"name": "forsythe-2016-ccm-image.png",
"type": "image/png",
"size": 401897,
"path": "Publication:forsythe-2016-ccm",
"url": "https://www.cg.tuwien.ac.at/research/publications/2016/forsythe-2016-ccm/forsythe-2016-ccm-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2016/forsythe-2016-ccm/forsythe-2016-ccm-image:thumb{{size}}.png"
},
{
"description": "null",
"filetitle": "paper",
"main_file": true,
"use_in_gallery": false,
"access": "public",
"name": "forsythe-2016-ccm-paper.pdf",
"type": "application/pdf",
"size": 3354515,
"path": "Publication:forsythe-2016-ccm",
"url": "https://www.cg.tuwien.ac.at/research/publications/2016/forsythe-2016-ccm/forsythe-2016-ccm-paper.pdf",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2016/forsythe-2016-ccm/forsythe-2016-ccm-paper:thumb{{size}}.png"
},
{
"description": null,
"filetitle": "slides",
"main_file": true,
"use_in_gallery": false,
"access": "public",
"name": "forsythe-2016-ccm-slides.pptx",
"type": "application/vnd.openxmlformats-officedocument.presentationml.presentation",
"size": 6295464,
"path": "Publication:forsythe-2016-ccm",
"url": "https://www.cg.tuwien.ac.at/research/publications/2016/forsythe-2016-ccm/forsythe-2016-ccm-slides.pptx",
"thumb_image_sizes": []
}
],
"projects_workgroups": [
"ShapeAcquisition"
],
"url": "https://www.cg.tuwien.ac.at/research/publications/2016/forsythe-2016-ccm/",
"__class": "Publication"
},
{
"id": "ohrhallinger-2016-sgp",
"type_id": "journalpaper",
"tu_id": null,
"repositum_id": null,
"title": "Curve Reconstruction with Many Fewer Samples",
"date": "2016",
"abstract": "We consider the problem of sampling points from a collection of smooth curves in the plane, such that the Crust family of proximity-based reconstruction algorithms can rebuild the curves. Reconstruction requires a dense sampling of local features, i.e., parts of the curve that are close in Euclidean distance but far apart geodesically.\nWe show that epsilon<0.47-sampling is sufficient for our proposed HNN-CRUST variant, improving upon the \nstate-of-the-art requirement of epsilon<1/3-sampling.\nThus we may reconstruct curves with many fewer samples.\nWe also present a new sampling scheme that reduces the required density even further than epsilon<0.47-sampling.\nWe achieve this by better controlling the spacing between geodesically consecutive points.\nOur novel sampling condition is based on the reach, the minimum local feature size along intervals between samples.\nThis is mathematically closer to the reconstruction density requirements, particularly near sharp-angled features.\nWe prove lower and upper bounds on reach rho-sampling density in terms of lfs epsilon-sampling and demonstrate that we typically reduce the required number of samples for reconstruction by more than half.\n",
"authors_et_al": false,
"substitute": null,
"main_image": {
"description": null,
"filetitle": "image",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 2000,
"image_height": 1200,
"name": "ohrhallinger-2016-sgp-image.png",
"type": "image/png",
"size": 54228,
"path": "Publication:ohrhallinger-2016-sgp",
"url": "https://www.cg.tuwien.ac.at/research/publications/2016/ohrhallinger-2016-sgp/ohrhallinger-2016-sgp-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2016/ohrhallinger-2016-sgp/ohrhallinger-2016-sgp-image:thumb{{size}}.png"
},
"sync_repositum_override": null,
"repositum_presentation_id": null,
"authors": [
948,
1363,
193
],
"date_from": "2016-06-20",
"date_to": "2016-06-24",
"event": "Symposium on Geometry Processing",
"issn": "1467-8659",
"journal": "Computer Graphics Forum",
"lecturer": [
948
],
"location": "Berlin, Germany",
"number": "5",
"pages_from": "167",
"pages_to": "176",
"volume": "35",
"research_areas": [
"Geometry"
],
"keywords": [
"sampling condition",
"curve reconstruction",
"curve sampling"
],
"weblinks": [
{
"href": "https://github.com/stefango74/hnn-crust-sgp16",
"caption": "Reproducibility Source Code",
"description": "This git repository contains all source code and instructions to reproduce the relevant figures and tables of the paper on an Ubuntu 16.04LTS system.",
"main_file": 1
},
{
"href": "http://dx.doi.org/10.1111/cgf.12973",
"caption": "doi",
"description": null,
"main_file": 0
}
],
"files": [
{
"description": null,
"filetitle": "image",
"main_file": false,
"use_in_gallery": true,
"access": "public",
"image_width": 2000,
"image_height": 1200,
"name": "ohrhallinger-2016-sgp-image.png",
"type": "image/png",
"size": 54228,
"path": "Publication:ohrhallinger-2016-sgp",
"url": "https://www.cg.tuwien.ac.at/research/publications/2016/ohrhallinger-2016-sgp/ohrhallinger-2016-sgp-image.png",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2016/ohrhallinger-2016-sgp/ohrhallinger-2016-sgp-image:thumb{{size}}.png"
},
{
"description": null,
"filetitle": "paper",
"main_file": true,
"use_in_gallery": false,
"access": "public",
"name": "ohrhallinger-2016-sgp-paper.pdf",
"type": "application/pdf",
"size": 969324,
"path": "Publication:ohrhallinger-2016-sgp",
"url": "https://www.cg.tuwien.ac.at/research/publications/2016/ohrhallinger-2016-sgp/ohrhallinger-2016-sgp-paper.pdf",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2016/ohrhallinger-2016-sgp/ohrhallinger-2016-sgp-paper:thumb{{size}}.png"
},
{
"description": "slides",
"filetitle": "slides",
"main_file": true,
"use_in_gallery": false,
"access": "public",
"name": "ohrhallinger-2016-sgp-slides.pdf",
"type": "application/pdf",
"size": 6164109,
"path": "Publication:ohrhallinger-2016-sgp",
"url": "https://www.cg.tuwien.ac.at/research/publications/2016/ohrhallinger-2016-sgp/ohrhallinger-2016-sgp-slides.pdf",
"thumb_image_sizes": [
16,
64,
100,
175,
300,
600
],
"thumb_url": "https://www.cg.tuwien.ac.at/research/publications/2016/ohrhallinger-2016-sgp/ohrhallinger-2016-sgp-slides:thumb{{size}}.png"
}
],
"projects_workgroups": [
"Harvest4D",
"ShapeAcquisition"
],
"url": "https://www.cg.tuwien.ac.at/research/publications/2016/ohrhallinger-2016-sgp/",
"__class": "Publication"
}
]