Sensor-Fusion for Smartphone Location Tracking Using Hybrid Multimodal Deep Neural Networks

Many engineered approaches have been proposed over the years for solving the hard problem of performing indoor localization using smartphone sensors. However, specialising these solutions for difficult edge cases remains challenging. Here we propose an end-to-end hybrid multimodal deep neural networ...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Xijia Wei, Zhiqiang Wei, Valentin Radu
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	indoor localization sensor fusion multimodal deep neural network multimodal sensing wifi fingerprinting pedestrian dead reckoning Chemical technology TP1-1185
Acceso en línea:	https://doaj.org/article/f5f97214bfc74fd999d152b1db17ef7f
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:f5f97214bfc74fd999d152b1db17ef7f
record_format	dspace
spelling	oai:doaj.org-article:f5f97214bfc74fd999d152b1db17ef7f2021-11-25T18:56:51ZSensor-Fusion for Smartphone Location Tracking Using Hybrid Multimodal Deep Neural Networks10.3390/s212274881424-8220https://doaj.org/article/f5f97214bfc74fd999d152b1db17ef7f2021-11-01T00:00:00Zhttps://www.mdpi.com/1424-8220/21/22/7488https://doaj.org/toc/1424-8220Many engineered approaches have been proposed over the years for solving the hard problem of performing indoor localization using smartphone sensors. However, specialising these solutions for difficult edge cases remains challenging. Here we propose an end-to-end hybrid multimodal deep neural network localization system, MM-Loc, relying on zero hand-engineered features, but learning automatically from data instead. This is achieved by using modality-specific neural networks to extract preliminary features from each sensing modality, which are then combined by cross-modality neural structures. We show that our choice of modality-specific neural architectures can estimate the location independently. But for better accuracy, a multimodal neural network that fuses the features of early modality-specific representations is a better proposition. Our proposed MM-Loc system is tested on cross-modality samples characterised by different sampling rate and data representation (inertial sensors, magnetic and WiFi signals), outperforming traditional approaches for location estimation. MM-Loc elegantly trains directly from data unlike conventional indoor positioning systems, which rely on human intuition.Xijia WeiZhiqiang WeiValentin RaduMDPI AGarticleindoor localizationsensor fusionmultimodal deep neural networkmultimodal sensingwifi fingerprintingpedestrian dead reckoningChemical technologyTP1-1185ENSensors, Vol 21, Iss 7488, p 7488 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	indoor localization sensor fusion multimodal deep neural network multimodal sensing wifi fingerprinting pedestrian dead reckoning Chemical technology TP1-1185
spellingShingle	indoor localization sensor fusion multimodal deep neural network multimodal sensing wifi fingerprinting pedestrian dead reckoning Chemical technology TP1-1185 Xijia Wei Zhiqiang Wei Valentin Radu Sensor-Fusion for Smartphone Location Tracking Using Hybrid Multimodal Deep Neural Networks
description	Many engineered approaches have been proposed over the years for solving the hard problem of performing indoor localization using smartphone sensors. However, specialising these solutions for difficult edge cases remains challenging. Here we propose an end-to-end hybrid multimodal deep neural network localization system, MM-Loc, relying on zero hand-engineered features, but learning automatically from data instead. This is achieved by using modality-specific neural networks to extract preliminary features from each sensing modality, which are then combined by cross-modality neural structures. We show that our choice of modality-specific neural architectures can estimate the location independently. But for better accuracy, a multimodal neural network that fuses the features of early modality-specific representations is a better proposition. Our proposed MM-Loc system is tested on cross-modality samples characterised by different sampling rate and data representation (inertial sensors, magnetic and WiFi signals), outperforming traditional approaches for location estimation. MM-Loc elegantly trains directly from data unlike conventional indoor positioning systems, which rely on human intuition.
format	article
author	Xijia Wei Zhiqiang Wei Valentin Radu
author_facet	Xijia Wei Zhiqiang Wei Valentin Radu
author_sort	Xijia Wei
title	Sensor-Fusion for Smartphone Location Tracking Using Hybrid Multimodal Deep Neural Networks
title_short	Sensor-Fusion for Smartphone Location Tracking Using Hybrid Multimodal Deep Neural Networks
title_full	Sensor-Fusion for Smartphone Location Tracking Using Hybrid Multimodal Deep Neural Networks
title_fullStr	Sensor-Fusion for Smartphone Location Tracking Using Hybrid Multimodal Deep Neural Networks
title_full_unstemmed	Sensor-Fusion for Smartphone Location Tracking Using Hybrid Multimodal Deep Neural Networks
title_sort	sensor-fusion for smartphone location tracking using hybrid multimodal deep neural networks
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/f5f97214bfc74fd999d152b1db17ef7f
work_keys_str_mv	AT xijiawei sensorfusionforsmartphonelocationtrackingusinghybridmultimodaldeepneuralnetworks AT zhiqiangwei sensorfusionforsmartphonelocationtrackingusinghybridmultimodaldeepneuralnetworks AT valentinradu sensorfusionforsmartphonelocationtrackingusinghybridmultimodaldeepneuralnetworks
_version_	1718410564103831552

Sensor-Fusion for Smartphone Location Tracking Using Hybrid Multimodal Deep Neural Networks

Ejemplares similares