7 research outputs found
Fast head profile estimation using curvature, derivatives and deep learning methods
Get PDFFast estimation of head profile and posture has applications across many disciplines, for example, it can be used in sleep apnoea screening and orthodontic examination or could support a suitable physiotherapy regime. Consequently, this thesis focuses on the investigation of methods to estimate head profile and posture efficiently and accurately, and results in the development and evaluation of datasets, features and deep learning models that can achieve this. Accordingly, this thesis initially investigated properties of contour curves that could act as effective features to train machine learning models. Features based on curvature and the first and second Gaussian derivatives were evaluated. These outperformed established features used in the literature to train a long short-term memory recurrent neural network and produced a significant speedup in execution time where pre-filtering of a sampled dataset was required. Following on from this, a new dataset of head profile contours was generated and annotated with anthropometric cranio-facial landmarks, and a novel method of automatically improving the accuracy of the landmark positions was developed using ideas based on the curvature of a plane curve. The features identified here were extracted from the new head profile contour dataset and used to train long short-term recurrent neural networks.
The best network, using Gaussian derivatives features achieved an accuracy of 91% and macro F1 score of 91%, an improvement of 51% and 71% respectively when compared with the un-processed contour feature. When using Gaussian derivative features, the network was able to regress landmarks accurately with mean absolute errors ranging from 0 to 5.3 pixels and standard deviations ranging from 0 to 6.9, respectively. End-to-end machine learning approaches, where a deep neural network learns the best features to use from the raw input data, were also investigated. Such an approach, using a one-dimensional temporal convolutional network was able to match previous classifiers in terms of accuracy and macro F1 score, and showed comparable regression abilities. However, this was at the expense of increased training times and increased inference times. This network was an order of magnitude slower when classifying and regressing contours
Re-Evaluation of RF Electromagnetic Communication in Underwater Sensor Networks
Get PDFMost underwater wireless networks use acoustic
waves as the transmission medium nowadays,
but the chances of getting much more out of
acoustic modems are quite remote. Optical links
are impractical for many underwater applications.
Given modern operational requirements
and digital communications technology, the time
is now ripe for re-evaluating the role of electromagnetic
signals in underwater environments.
The research presented in this article is motivated
by the limitations of current and established
wireless underwater techniques, as well as the
potential that electromagnetic waves can offer to
underwater applications. A case study is presented
that uses electromagnetic technology in a
small-scale underwater wireless sensor network.
The results demonstrate the likely effectiveness
of the designated network
TDMA frame design for a prototype underwater RF
Get PDFVery low frequency electromagnetic communication system is used in a small scale underwater
wireless sensor network for coastal monitoring purposes, as recent research has
demonstrated distinct advantages of radio waves compared to acoustic and optical waves
in shallow water conditions. This paper describes the detailed TDMA and packet design
process for the prototype sensor system. The lightweight protocol is time division based
in order to fit the unique characteristics and specifications of the network. Evaluations
are based on initial beach trial as well as modeling and simulations
TDMA frame design for a prototype underwater RF communication network
Get PDFThis document is the Accepted Manuscript version of the following article: Xianhui Che, Ian Wells, Gordon Dickers, and Paul Kear, ‘TDMA frame design for a prototype underwater RF communication network’, Ad Hoc Networks, Vol. 10 (3): 317-327, first available online 23 July 2011. The version of record is available online at doi: http://dx.doi.org/10.1016/j.adhoc.2011.07.002 © 2011 Elsevier B. V. All rights reserved.Very low frequency electromagnetic communication system is used in a small scale underwater wireless sensor network for coastal monitoring purposes, as recent research has demonstrated distinct advantages of radio waves compared to acoustic and optical waves in shallow water conditions. This paper describes the detailed TDMA and packet design process for the prototype sensor system. The lightweight protocol is time division based in order to fit the unique characteristics and specifications of the network. Evaluations are based on initial beach trial as well as modeling and simulations.Peer reviewe
A Static Multi-hop Underwater Wireless Sensor Network Using RF Electromagnetic Communications
No full textMost underwater sensor networks choose acoustics as the medium for wireless transmission. However, electromagnetic waves also offer great merits for transmission in special underwater environment. A small scale wireless sensor network is deployed using electromagnetic waves with a multi-hop static topology under shallow water conditions where there is a high level of sediment and aeration in the water column. Data delivery is scheduled via daily cycles of sleeping and waking up to transmit. Due to the unique features of the network, ad-hoc on-demand distance vector (AODV) is chosen as the routing protocol. Modeling and simulations are conducted to evaluate network performance in terms of failure tolerance, congestion handling, and optimal grid arrangements. The results demonstrate the likely effectiveness of the designated network for this and similar scenarios
Failure tolerance analysis of a small scale underwater sensor network with RF electromagnetic communications
No full textA small scale wireless sensor network was designed using electromagnetic technology in order to provide an underwater coastal monitoring service. The sensor nodes were deployed manually, which formed a multi-hop static topology for the network. Data delivery was scheduled via daily cycles of sleeping and waking-up. Due to the unique features of the network, Ad-hoc On-demand Distance Vector was chosen as the routing protocol. Modeling and simulations were conducted to evaluate network performance in terms of failure recovery and network re-convergence. The results demonstrated a certain level of failure tolerance of the designated network for this and similar scenarios
