Predicting consumers' locations in dynamic environments via 3D sensor-based tracking

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

1 Citation (Scopus)

Abstract

Brick-and-mortar stores for non-food items allow customers to quickly try items and take them home, but lack certain convenient features of online shopping, such as personalised offers and recommendations for items of possible interest. Mobile in-store shopping applications would allow to combine these advantages if they would derive current customer needs from customer activities. A natural way to infer interests of shops' visitors is to analyse their motion and places where they stop. This paper presents a low-cost depth sensor -- based people tracking system and a method to predict future customer locations, developed for environments where items are frequently re-located and customer routes change accordingly. The tracking system employs adaptive background modelling approach, allowing to quickly distinguish between moving humans and re-located objects. Similarity-based location predictions use fairly small datasets of recent tracks of other customers and allow predicting locations of future stops very soon after monitoring starts: after just a few minutes. Therefore this approach also provides for imperfect tracking, e.g., due to occlusions. In the tests with the data, acquired in a real clothing and cosmetics department during 50 days, future places of customers' interest were predicted with an average accuracy 60% and an average distance error half a metre.
Original languageEnglish
Title of host publicationProceedings
Subtitle of host publicationEighth International Conference on Next Generation Mobile Apps, Services and Technologies, NGMAST 2014
PublisherIEEE Institute of Electrical and Electronic Engineers
Pages100-105
ISBN (Electronic)978-1-4799-5073-7
ISBN (Print)978-1-4799-5072-0
DOIs
Publication statusPublished - 2014
MoE publication typeA4 Article in a conference publication
Event8th International Conference on Next Generation Mobile Apps, Services and Technologies, NGMAST 2014 - Oxford, United Kingdom
Duration: 10 Sep 201412 Sep 2014
Conference number: 8th

Conference

Conference8th International Conference on Next Generation Mobile Apps, Services and Technologies, NGMAST 2014
Abbreviated titleNGMAST 2014
CountryUnited Kingdom
CityOxford
Period10/09/1412/09/14

Fingerprint

Sensors
Cosmetics
Adaptive systems
Brick
Mortar
Monitoring
Costs

Keywords

  • consumr behaviour
  • image sensor
  • mobile computing
  • object tracking
  • retail data processing

Cite this

Vildjiounaite, E., Mäkelä, S-M., Järvinen, S., Keränen, T., & Kyllönen, V. (2014). Predicting consumers' locations in dynamic environments via 3D sensor-based tracking. In Proceedings: Eighth International Conference on Next Generation Mobile Apps, Services and Technologies, NGMAST 2014 (pp. 100-105). IEEE Institute of Electrical and Electronic Engineers . https://doi.org/10.1109/NGMAST.2014.11
Vildjiounaite, Elena ; Mäkelä, Satu-Marja ; Järvinen, Sari ; Keränen, Tommi ; Kyllönen, Vesa. / Predicting consumers' locations in dynamic environments via 3D sensor-based tracking. Proceedings: Eighth International Conference on Next Generation Mobile Apps, Services and Technologies, NGMAST 2014. IEEE Institute of Electrical and Electronic Engineers , 2014. pp. 100-105
@inproceedings{f528a9ff157b4586beee5c46bcbc1a6e,
title = "Predicting consumers' locations in dynamic environments via 3D sensor-based tracking",
abstract = "Brick-and-mortar stores for non-food items allow customers to quickly try items and take them home, but lack certain convenient features of online shopping, such as personalised offers and recommendations for items of possible interest. Mobile in-store shopping applications would allow to combine these advantages if they would derive current customer needs from customer activities. A natural way to infer interests of shops' visitors is to analyse their motion and places where they stop. This paper presents a low-cost depth sensor -- based people tracking system and a method to predict future customer locations, developed for environments where items are frequently re-located and customer routes change accordingly. The tracking system employs adaptive background modelling approach, allowing to quickly distinguish between moving humans and re-located objects. Similarity-based location predictions use fairly small datasets of recent tracks of other customers and allow predicting locations of future stops very soon after monitoring starts: after just a few minutes. Therefore this approach also provides for imperfect tracking, e.g., due to occlusions. In the tests with the data, acquired in a real clothing and cosmetics department during 50 days, future places of customers' interest were predicted with an average accuracy 60{\%} and an average distance error half a metre.",
keywords = "consumr behaviour, image sensor, mobile computing, object tracking, retail data processing",
author = "Elena Vildjiounaite and Satu-Marja M{\"a}kel{\"a} and Sari J{\"a}rvinen and Tommi Ker{\"a}nen and Vesa Kyll{\"o}nen",
year = "2014",
doi = "10.1109/NGMAST.2014.11",
language = "English",
isbn = "978-1-4799-5072-0",
pages = "100--105",
booktitle = "Proceedings",
publisher = "IEEE Institute of Electrical and Electronic Engineers",
address = "United States",

}

Vildjiounaite, E, Mäkelä, S-M, Järvinen, S, Keränen, T & Kyllönen, V 2014, Predicting consumers' locations in dynamic environments via 3D sensor-based tracking. in Proceedings: Eighth International Conference on Next Generation Mobile Apps, Services and Technologies, NGMAST 2014. IEEE Institute of Electrical and Electronic Engineers , pp. 100-105, 8th International Conference on Next Generation Mobile Apps, Services and Technologies, NGMAST 2014, Oxford, United Kingdom, 10/09/14. https://doi.org/10.1109/NGMAST.2014.11

Predicting consumers' locations in dynamic environments via 3D sensor-based tracking. / Vildjiounaite, Elena; Mäkelä, Satu-Marja; Järvinen, Sari; Keränen, Tommi; Kyllönen, Vesa.

Proceedings: Eighth International Conference on Next Generation Mobile Apps, Services and Technologies, NGMAST 2014. IEEE Institute of Electrical and Electronic Engineers , 2014. p. 100-105.

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

TY - GEN

T1 - Predicting consumers' locations in dynamic environments via 3D sensor-based tracking

AU - Vildjiounaite, Elena

AU - Mäkelä, Satu-Marja

AU - Järvinen, Sari

AU - Keränen, Tommi

AU - Kyllönen, Vesa

PY - 2014

Y1 - 2014

N2 - Brick-and-mortar stores for non-food items allow customers to quickly try items and take them home, but lack certain convenient features of online shopping, such as personalised offers and recommendations for items of possible interest. Mobile in-store shopping applications would allow to combine these advantages if they would derive current customer needs from customer activities. A natural way to infer interests of shops' visitors is to analyse their motion and places where they stop. This paper presents a low-cost depth sensor -- based people tracking system and a method to predict future customer locations, developed for environments where items are frequently re-located and customer routes change accordingly. The tracking system employs adaptive background modelling approach, allowing to quickly distinguish between moving humans and re-located objects. Similarity-based location predictions use fairly small datasets of recent tracks of other customers and allow predicting locations of future stops very soon after monitoring starts: after just a few minutes. Therefore this approach also provides for imperfect tracking, e.g., due to occlusions. In the tests with the data, acquired in a real clothing and cosmetics department during 50 days, future places of customers' interest were predicted with an average accuracy 60% and an average distance error half a metre.

AB - Brick-and-mortar stores for non-food items allow customers to quickly try items and take them home, but lack certain convenient features of online shopping, such as personalised offers and recommendations for items of possible interest. Mobile in-store shopping applications would allow to combine these advantages if they would derive current customer needs from customer activities. A natural way to infer interests of shops' visitors is to analyse their motion and places where they stop. This paper presents a low-cost depth sensor -- based people tracking system and a method to predict future customer locations, developed for environments where items are frequently re-located and customer routes change accordingly. The tracking system employs adaptive background modelling approach, allowing to quickly distinguish between moving humans and re-located objects. Similarity-based location predictions use fairly small datasets of recent tracks of other customers and allow predicting locations of future stops very soon after monitoring starts: after just a few minutes. Therefore this approach also provides for imperfect tracking, e.g., due to occlusions. In the tests with the data, acquired in a real clothing and cosmetics department during 50 days, future places of customers' interest were predicted with an average accuracy 60% and an average distance error half a metre.

KW - consumr behaviour

KW - image sensor

KW - mobile computing

KW - object tracking

KW - retail data processing

U2 - 10.1109/NGMAST.2014.11

DO - 10.1109/NGMAST.2014.11

M3 - Conference article in proceedings

SN - 978-1-4799-5072-0

SP - 100

EP - 105

BT - Proceedings

PB - IEEE Institute of Electrical and Electronic Engineers

ER -

Vildjiounaite E, Mäkelä S-M, Järvinen S, Keränen T, Kyllönen V. Predicting consumers' locations in dynamic environments via 3D sensor-based tracking. In Proceedings: Eighth International Conference on Next Generation Mobile Apps, Services and Technologies, NGMAST 2014. IEEE Institute of Electrical and Electronic Engineers . 2014. p. 100-105 https://doi.org/10.1109/NGMAST.2014.11