Memory Efficient 3D Integral Volumes

Martin Urschler, Alexander Bornik, Michael Donoser

Publikation: Beitrag in Buch/Bericht/KonferenzbandBeitrag in einem KonferenzbandForschungBegutachtung

Abstract

Integral image data structures are very useful in computer vision applications that involve machine learning approaches based on ensembles of weak learners. The weak learners often are simply several regional sums of intensities subtracted from each other. In this work we present a memory efficient integral volume data structure, that allows reduction of required RAM storage size in such a supervised learning framework using 3D training data. We evaluate our proposed data structure in terms of the tradeoff between computational effort and storage, and show an application for 3D object detection of liver CT data.
Originalspracheenglisch
TitelIEEE International Conference on Computer Vision Workshops (ICCVW)
UntertitelBig Data in 3D Computer Vision
Herausgeber (Verlag)Institute of Electrical and Electronics Engineers
Seiten722-729
DOIs
PublikationsstatusVeröffentlicht - 2013

Fingerprint

Data structures
Data storage equipment
Supervised learning
Random access storage
Liver
Computer vision
Learning systems
Object detection

Fields of Expertise

  • Information, Communication & Computing

Dies zitieren

Urschler, M., Bornik, A., & Donoser, M. (2013). Memory Efficient 3D Integral Volumes. in IEEE International Conference on Computer Vision Workshops (ICCVW): Big Data in 3D Computer Vision (S. 722-729). Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/ICCVW.2013.99

Memory Efficient 3D Integral Volumes. / Urschler, Martin; Bornik, Alexander; Donoser, Michael.

IEEE International Conference on Computer Vision Workshops (ICCVW): Big Data in 3D Computer Vision. Institute of Electrical and Electronics Engineers, 2013. S. 722-729.

Publikation: Beitrag in Buch/Bericht/KonferenzbandBeitrag in einem KonferenzbandForschungBegutachtung

Urschler, M, Bornik, A & Donoser, M 2013, Memory Efficient 3D Integral Volumes. in IEEE International Conference on Computer Vision Workshops (ICCVW): Big Data in 3D Computer Vision. Institute of Electrical and Electronics Engineers, S. 722-729. https://doi.org/10.1109/ICCVW.2013.99
Urschler M, Bornik A, Donoser M. Memory Efficient 3D Integral Volumes. in IEEE International Conference on Computer Vision Workshops (ICCVW): Big Data in 3D Computer Vision. Institute of Electrical and Electronics Engineers. 2013. S. 722-729 https://doi.org/10.1109/ICCVW.2013.99
Urschler, Martin ; Bornik, Alexander ; Donoser, Michael. / Memory Efficient 3D Integral Volumes. IEEE International Conference on Computer Vision Workshops (ICCVW): Big Data in 3D Computer Vision. Institute of Electrical and Electronics Engineers, 2013. S. 722-729
@inproceedings{f5721c575a4b4c4a83a529c11320a9e4,
title = "Memory Efficient 3D Integral Volumes",
abstract = "Integral image data structures are very useful in computer vision applications that involve machine learning approaches based on ensembles of weak learners. The weak learners often are simply several regional sums of intensities subtracted from each other. In this work we present a memory efficient integral volume data structure, that allows reduction of required RAM storage size in such a supervised learning framework using 3D training data. We evaluate our proposed data structure in terms of the tradeoff between computational effort and storage, and show an application for 3D object detection of liver CT data.",
author = "Martin Urschler and Alexander Bornik and Michael Donoser",
year = "2013",
doi = "10.1109/ICCVW.2013.99",
language = "English",
pages = "722--729",
booktitle = "IEEE International Conference on Computer Vision Workshops (ICCVW)",
publisher = "Institute of Electrical and Electronics Engineers",
address = "United States",

}

TY - GEN

T1 - Memory Efficient 3D Integral Volumes

AU - Urschler, Martin

AU - Bornik, Alexander

AU - Donoser, Michael

PY - 2013

Y1 - 2013

N2 - Integral image data structures are very useful in computer vision applications that involve machine learning approaches based on ensembles of weak learners. The weak learners often are simply several regional sums of intensities subtracted from each other. In this work we present a memory efficient integral volume data structure, that allows reduction of required RAM storage size in such a supervised learning framework using 3D training data. We evaluate our proposed data structure in terms of the tradeoff between computational effort and storage, and show an application for 3D object detection of liver CT data.

AB - Integral image data structures are very useful in computer vision applications that involve machine learning approaches based on ensembles of weak learners. The weak learners often are simply several regional sums of intensities subtracted from each other. In this work we present a memory efficient integral volume data structure, that allows reduction of required RAM storage size in such a supervised learning framework using 3D training data. We evaluate our proposed data structure in terms of the tradeoff between computational effort and storage, and show an application for 3D object detection of liver CT data.

U2 - 10.1109/ICCVW.2013.99

DO - 10.1109/ICCVW.2013.99

M3 - Conference contribution

SP - 722

EP - 729

BT - IEEE International Conference on Computer Vision Workshops (ICCVW)

PB - Institute of Electrical and Electronics Engineers

ER -