Declutter and resample: Towards parameter free denoising

Mickaël Buchet, Tamal K. Dey, Jiayuan Wang, Yusu Wang

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Abstract

In many data analysis applications the following scenario is commonplace: we are given a point set that is supposed to sample a hidden ground truth K in a metric space, but it got corrupted with noise so that some of the data points lie far away from K creating outliers also termed as ambient noise. One of the main goals of denoising algorithms is to eliminate such noise so that the curated data lie within a bounded Hausdorff distance of K. Popular denoising approaches such as deconvolution and thresholding often require the user to set several parameters and/or to choose an appropriate noise model while guaranteeing only asymptotic convergence. Our goal is to lighten this burden as much as possible while ensuring theoretical guarantees in all cases. Specifically, first, we propose a simple denoising algorithm that requires only a single parameter but provides a theoretical guarantee on the quality of the output on general input points. We argue that this single parameter cannot be avoided. We next present a simple algorithm that avoids even this parameter by paying for it with a slight strengthening of the sampling condition on the input points which is not unrealistic. We also provide some preliminary empirical evidence that our algorithms are effective in practice.

Original languageEnglish
Title of host publication33rd International Symposium on Computational Geometry, SoCG 2017
PublisherSchloss Dagstuhl, Leibniz-Zentrum fü Informatik GmbH
Pages231-2316
Number of pages2086
Volume77
ISBN (Electronic)9783959770385
DOIs
Publication statusPublished - 1 Jun 2017
Externally publishedYes
Event33rd International Symposium on Computational Geometry, SoCG 2017 - Brisbane, Australia
Duration: 4 Jul 20177 Jul 2017

Conference

Conference33rd International Symposium on Computational Geometry, SoCG 2017
CountryAustralia
CityBrisbane
Period4/07/177/07/17

Fingerprint

Deconvolution
Sampling

Keywords

  • Compact sets
  • Denoising
  • K-distance
  • Parameter free

ASJC Scopus subject areas

  • Software

Cite this

Buchet, M., Dey, T. K., Wang, J., & Wang, Y. (2017). Declutter and resample: Towards parameter free denoising. In 33rd International Symposium on Computational Geometry, SoCG 2017 (Vol. 77, pp. 231-2316). Schloss Dagstuhl, Leibniz-Zentrum fü Informatik GmbH. https://doi.org/10.4230/LIPIcs.SoCG.2017.23

Declutter and resample : Towards parameter free denoising. / Buchet, Mickaël; Dey, Tamal K.; Wang, Jiayuan; Wang, Yusu.

33rd International Symposium on Computational Geometry, SoCG 2017. Vol. 77 Schloss Dagstuhl, Leibniz-Zentrum fü Informatik GmbH, 2017. p. 231-2316.

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Buchet, M, Dey, TK, Wang, J & Wang, Y 2017, Declutter and resample: Towards parameter free denoising. in 33rd International Symposium on Computational Geometry, SoCG 2017. vol. 77, Schloss Dagstuhl, Leibniz-Zentrum fü Informatik GmbH, pp. 231-2316, 33rd International Symposium on Computational Geometry, SoCG 2017, Brisbane, Australia, 4/07/17. https://doi.org/10.4230/LIPIcs.SoCG.2017.23
Buchet M, Dey TK, Wang J, Wang Y. Declutter and resample: Towards parameter free denoising. In 33rd International Symposium on Computational Geometry, SoCG 2017. Vol. 77. Schloss Dagstuhl, Leibniz-Zentrum fü Informatik GmbH. 2017. p. 231-2316 https://doi.org/10.4230/LIPIcs.SoCG.2017.23
Buchet, Mickaël ; Dey, Tamal K. ; Wang, Jiayuan ; Wang, Yusu. / Declutter and resample : Towards parameter free denoising. 33rd International Symposium on Computational Geometry, SoCG 2017. Vol. 77 Schloss Dagstuhl, Leibniz-Zentrum fü Informatik GmbH, 2017. pp. 231-2316
@inproceedings{41b8b651eed946a4993e29eec0b0ffbf,
title = "Declutter and resample: Towards parameter free denoising",
abstract = "In many data analysis applications the following scenario is commonplace: we are given a point set that is supposed to sample a hidden ground truth K in a metric space, but it got corrupted with noise so that some of the data points lie far away from K creating outliers also termed as ambient noise. One of the main goals of denoising algorithms is to eliminate such noise so that the curated data lie within a bounded Hausdorff distance of K. Popular denoising approaches such as deconvolution and thresholding often require the user to set several parameters and/or to choose an appropriate noise model while guaranteeing only asymptotic convergence. Our goal is to lighten this burden as much as possible while ensuring theoretical guarantees in all cases. Specifically, first, we propose a simple denoising algorithm that requires only a single parameter but provides a theoretical guarantee on the quality of the output on general input points. We argue that this single parameter cannot be avoided. We next present a simple algorithm that avoids even this parameter by paying for it with a slight strengthening of the sampling condition on the input points which is not unrealistic. We also provide some preliminary empirical evidence that our algorithms are effective in practice.",
keywords = "Compact sets, Denoising, K-distance, Parameter free",
author = "Micka{\"e}l Buchet and Dey, {Tamal K.} and Jiayuan Wang and Yusu Wang",
year = "2017",
month = "6",
day = "1",
doi = "10.4230/LIPIcs.SoCG.2017.23",
language = "English",
volume = "77",
pages = "231--2316",
booktitle = "33rd International Symposium on Computational Geometry, SoCG 2017",
publisher = "Schloss Dagstuhl, Leibniz-Zentrum f{\"u} Informatik GmbH",
address = "Germany",

}

TY - GEN

T1 - Declutter and resample

T2 - Towards parameter free denoising

AU - Buchet, Mickaël

AU - Dey, Tamal K.

AU - Wang, Jiayuan

AU - Wang, Yusu

PY - 2017/6/1

Y1 - 2017/6/1

N2 - In many data analysis applications the following scenario is commonplace: we are given a point set that is supposed to sample a hidden ground truth K in a metric space, but it got corrupted with noise so that some of the data points lie far away from K creating outliers also termed as ambient noise. One of the main goals of denoising algorithms is to eliminate such noise so that the curated data lie within a bounded Hausdorff distance of K. Popular denoising approaches such as deconvolution and thresholding often require the user to set several parameters and/or to choose an appropriate noise model while guaranteeing only asymptotic convergence. Our goal is to lighten this burden as much as possible while ensuring theoretical guarantees in all cases. Specifically, first, we propose a simple denoising algorithm that requires only a single parameter but provides a theoretical guarantee on the quality of the output on general input points. We argue that this single parameter cannot be avoided. We next present a simple algorithm that avoids even this parameter by paying for it with a slight strengthening of the sampling condition on the input points which is not unrealistic. We also provide some preliminary empirical evidence that our algorithms are effective in practice.

AB - In many data analysis applications the following scenario is commonplace: we are given a point set that is supposed to sample a hidden ground truth K in a metric space, but it got corrupted with noise so that some of the data points lie far away from K creating outliers also termed as ambient noise. One of the main goals of denoising algorithms is to eliminate such noise so that the curated data lie within a bounded Hausdorff distance of K. Popular denoising approaches such as deconvolution and thresholding often require the user to set several parameters and/or to choose an appropriate noise model while guaranteeing only asymptotic convergence. Our goal is to lighten this burden as much as possible while ensuring theoretical guarantees in all cases. Specifically, first, we propose a simple denoising algorithm that requires only a single parameter but provides a theoretical guarantee on the quality of the output on general input points. We argue that this single parameter cannot be avoided. We next present a simple algorithm that avoids even this parameter by paying for it with a slight strengthening of the sampling condition on the input points which is not unrealistic. We also provide some preliminary empirical evidence that our algorithms are effective in practice.

KW - Compact sets

KW - Denoising

KW - K-distance

KW - Parameter free

UR - http://www.scopus.com/inward/record.url?scp=85029930392&partnerID=8YFLogxK

U2 - 10.4230/LIPIcs.SoCG.2017.23

DO - 10.4230/LIPIcs.SoCG.2017.23

M3 - Conference contribution

VL - 77

SP - 231

EP - 2316

BT - 33rd International Symposium on Computational Geometry, SoCG 2017

PB - Schloss Dagstuhl, Leibniz-Zentrum fü Informatik GmbH

ER -