Total variation regularization for training of indoor location fingerprints

Duc A. Tran, Phong Truong

Research output: Chapter in Book/Report/Conference proceedingConference contribution

13 Citations (Scopus)

Abstract

Location fingerprinting is a common approach to indoor localization. For good accuracy, the training set of sample fingerprints, each mapping a fingerprint to a location, should be sufficiently large to be well-representative of the environment in terms of both spatial coverage and temporal coverage. Unfortunately, the task of collecting these samples can be tedious and labor-intensive because one must label each location that is being surveyed. On the other hand, fingerprints without location information are abundant and can easily be collected and so recent studies have tried to capitalize on these unlabeled fingerprints to improve the training set. The paper investigates how this goal can be achieved via graph regularization based on Total Variation (TV). TV is highly effective for semi-supervised learning in image processing but it is not clear whether its success can be transferred to indoor location fingerprinting.

Original languageEnglish (US)
Title of host publicationMiSeNet 2013 - Proceedings of the 2nd ACM Annual International Workshop on Mission-Oriented Wireless Sensor Networking
Pages27-32
Number of pages6
DOIs
StatePublished - Nov 13 2013
Event2nd ACM Annual International Workshop on Mission-Oriented Wireless Sensor Networking, MiSeNet 2013 - Miami, FL, United States
Duration: Oct 4 2013Oct 4 2013

Publication series

NameProceedings of the Annual International Conference on Mobile Computing and Networking, MOBICOM

Conference

Conference2nd ACM Annual International Workshop on Mission-Oriented Wireless Sensor Networking, MiSeNet 2013
CountryUnited States
CityMiami, FL
Period10/4/1310/4/13

Fingerprint

Supervised learning
Labels
Image processing
Personnel

All Science Journal Classification (ASJC) codes

  • Software
  • Hardware and Architecture
  • Computer Networks and Communications

Cite this

Tran, D. A., & Truong, P. (2013). Total variation regularization for training of indoor location fingerprints. In MiSeNet 2013 - Proceedings of the 2nd ACM Annual International Workshop on Mission-Oriented Wireless Sensor Networking (pp. 27-32). (Proceedings of the Annual International Conference on Mobile Computing and Networking, MOBICOM). https://doi.org/10.1145/2509338.2509346
Tran, Duc A. ; Truong, Phong. / Total variation regularization for training of indoor location fingerprints. MiSeNet 2013 - Proceedings of the 2nd ACM Annual International Workshop on Mission-Oriented Wireless Sensor Networking. 2013. pp. 27-32 (Proceedings of the Annual International Conference on Mobile Computing and Networking, MOBICOM).
@inproceedings{c390c145e193461698c844bb5d0d39ed,
title = "Total variation regularization for training of indoor location fingerprints",
abstract = "Location fingerprinting is a common approach to indoor localization. For good accuracy, the training set of sample fingerprints, each mapping a fingerprint to a location, should be sufficiently large to be well-representative of the environment in terms of both spatial coverage and temporal coverage. Unfortunately, the task of collecting these samples can be tedious and labor-intensive because one must label each location that is being surveyed. On the other hand, fingerprints without location information are abundant and can easily be collected and so recent studies have tried to capitalize on these unlabeled fingerprints to improve the training set. The paper investigates how this goal can be achieved via graph regularization based on Total Variation (TV). TV is highly effective for semi-supervised learning in image processing but it is not clear whether its success can be transferred to indoor location fingerprinting.",
author = "Tran, {Duc A.} and Phong Truong",
year = "2013",
month = "11",
day = "13",
doi = "10.1145/2509338.2509346",
language = "English (US)",
isbn = "9781450323673",
series = "Proceedings of the Annual International Conference on Mobile Computing and Networking, MOBICOM",
pages = "27--32",
booktitle = "MiSeNet 2013 - Proceedings of the 2nd ACM Annual International Workshop on Mission-Oriented Wireless Sensor Networking",

}

Tran, DA & Truong, P 2013, Total variation regularization for training of indoor location fingerprints. in MiSeNet 2013 - Proceedings of the 2nd ACM Annual International Workshop on Mission-Oriented Wireless Sensor Networking. Proceedings of the Annual International Conference on Mobile Computing and Networking, MOBICOM, pp. 27-32, 2nd ACM Annual International Workshop on Mission-Oriented Wireless Sensor Networking, MiSeNet 2013, Miami, FL, United States, 10/4/13. https://doi.org/10.1145/2509338.2509346

Total variation regularization for training of indoor location fingerprints. / Tran, Duc A.; Truong, Phong.

MiSeNet 2013 - Proceedings of the 2nd ACM Annual International Workshop on Mission-Oriented Wireless Sensor Networking. 2013. p. 27-32 (Proceedings of the Annual International Conference on Mobile Computing and Networking, MOBICOM).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Total variation regularization for training of indoor location fingerprints

AU - Tran, Duc A.

AU - Truong, Phong

PY - 2013/11/13

Y1 - 2013/11/13

N2 - Location fingerprinting is a common approach to indoor localization. For good accuracy, the training set of sample fingerprints, each mapping a fingerprint to a location, should be sufficiently large to be well-representative of the environment in terms of both spatial coverage and temporal coverage. Unfortunately, the task of collecting these samples can be tedious and labor-intensive because one must label each location that is being surveyed. On the other hand, fingerprints without location information are abundant and can easily be collected and so recent studies have tried to capitalize on these unlabeled fingerprints to improve the training set. The paper investigates how this goal can be achieved via graph regularization based on Total Variation (TV). TV is highly effective for semi-supervised learning in image processing but it is not clear whether its success can be transferred to indoor location fingerprinting.

AB - Location fingerprinting is a common approach to indoor localization. For good accuracy, the training set of sample fingerprints, each mapping a fingerprint to a location, should be sufficiently large to be well-representative of the environment in terms of both spatial coverage and temporal coverage. Unfortunately, the task of collecting these samples can be tedious and labor-intensive because one must label each location that is being surveyed. On the other hand, fingerprints without location information are abundant and can easily be collected and so recent studies have tried to capitalize on these unlabeled fingerprints to improve the training set. The paper investigates how this goal can be achieved via graph regularization based on Total Variation (TV). TV is highly effective for semi-supervised learning in image processing but it is not clear whether its success can be transferred to indoor location fingerprinting.

UR - http://www.scopus.com/inward/record.url?scp=84887216422&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84887216422&partnerID=8YFLogxK

U2 - 10.1145/2509338.2509346

DO - 10.1145/2509338.2509346

M3 - Conference contribution

AN - SCOPUS:84887216422

SN - 9781450323673

T3 - Proceedings of the Annual International Conference on Mobile Computing and Networking, MOBICOM

SP - 27

EP - 32

BT - MiSeNet 2013 - Proceedings of the 2nd ACM Annual International Workshop on Mission-Oriented Wireless Sensor Networking

ER -

Tran DA, Truong P. Total variation regularization for training of indoor location fingerprints. In MiSeNet 2013 - Proceedings of the 2nd ACM Annual International Workshop on Mission-Oriented Wireless Sensor Networking. 2013. p. 27-32. (Proceedings of the Annual International Conference on Mobile Computing and Networking, MOBICOM). https://doi.org/10.1145/2509338.2509346