Skip to main content

Rajib Rana (rajibrana.org)

University of Southern Queensland, Centre for Health Sciences Research, Research Scientist

The University of Queensland, Australia, School of ITEE, Adjunct

Followers

28

Following

10

Co-authors

9

Public Views

My vision is to be a pioneer in the field of information theory into practical applications.

Skills:-

- Strong technical, analytical, problem solving and research skills obtained through diverse projects at different institutions.
- Demonstrated leadership and organizational skills as taken leading professional roles such as supervisor for honors and postgraduate students, and group leader in group-based projects.
- Exceptional communication, interpersonal and written skills demonstrated as a presenter at international conferences, through collaborative work with international institutions, and international peer-reviewed publications.
- Proven track record of research with 30 internationally peer reviewed conference and journal publications.
- Practical hands-on experience of analysis, design and development of desktop and embedded hardware/software systems using complex algorithm design, software architecture design and programming.
- 3+ years experience in teaching at university level.
- 7+ years experience in Python, Python S60, Java and Matlab.

Technical Specialties: C/C++/Java/Perl/Python, MATLAB programming, Linux/Windows scripting.

Research Specialties: Compressive Sensing, Applied Machine Learning, Optimization.

Academic achievements/grants:-

-2014-2015: 100k Research grant from Queensland Health.
-Short listed for Australian Health Innovation Fund (AusHSI) - 2013
-Recipient-UNSW writing postdoctoral fellowship - 2011
-Nomination for BEST thesis award - 2011
-University International postgraduate research award - 2007
-President's GOLD medal (BSc.) - 2005
-Prime Minister's GOLD medal (BSc.) - 2005

Collaborators:-

UCLA,
ORCATECH,
Portland State University,
QIMR,
Queensland Health,
Black Dog Institute,
Supervisors: Chun Tung Chou

less

The University of New South Wales

Nirupama Bulusu

Nottingham Trent University

Roma Tre University, Rome, Italy

University of California, Berkeley

Teerayut Horanont

The University of Tokyo

The University of New South Wales

Emiliano Miluzzo

Dartmouth College

InterestsView All (6)

Uploads

Papers by Rajib Rana (rajibrana.org)

by Nirupama Bulusu, Chun Chou, and Rajib Rana (rajibrana.org)

Proceedings of the 9th ACM/IEEE International Conference on Information Processing in Sensor Networks - IPSN '10, 2010

A noise map facilitates monitoring of environmental noise pollution in urban areas. It can raise ... more A noise map facilitates monitoring of environmental noise pollution in urban areas. It can raise citizen awareness of noise pollution levels, and aid in the development of mitigation strategies to cope with the adverse effects. However, state-of-the-art techniques for rendering noise maps in urban areas are expensive and rarely updated (months or even years), as they rely on population and traffic models rather than on real data. Participatory urban sensing can be leveraged to create an open and inexpensive platform for rendering up-to-date noise maps.

Nonuniform compressive sensing for heterogeneous wireless sensor networks

by Chun Chou and Rajib Rana (rajibrana.org)

In this paper, we consider the problem of using wireless sensor networks (WSNs) to measure the te... more In this paper, we consider the problem of using wireless sensor networks (WSNs) to measure the temporal-spatial profile of some physical phenomena. We base our work on two observations. First, most physical phenomena are compressible in some transform domain basis. Second, most WSNs have some form of heterogeneity. Given these two observations, we propose a nonuniform compressive sensing method to improve the performance of WSNs by exploiting both compressibility and heterogeneity. We apply our proposed method to real WSN data sets. We find that our method can provide a more accurate temporal-spatial profile for a given energy budget compared with other sampling methods.

Non-uniform compressive sensing in wireless sensor networks: Feasibility and application

by Chun Chou, Rajib Rana (rajibrana.org), and W. Hu

In this paper, we consider the problem of using wireless sensor networks (WSNs) to measure the te... more In this paper, we consider the problem of using wireless sensor networks (WSNs) to measure the temporal-spatial profile of some physical phenomena. We base our work on two observations. Firstly, most physical phenomena are compressible in some transform domain basis. Secondly, most WSNs have some form of heterogeneity. Given these two observations, we propose a non-uniform compressive sensing method to improve the performance of WSNs by exploiting both compressibility and heterogeneity. We apply our proposed method to a real WSN data set. We find that our method can provide a more accurate temporal-spatial profile for a given energy budget compared with other sampling methods.

Ear-Phone: An End-to-End Participatory Urban Noise Mapping System

by Chun Chou, Nirupama Bulusu, and Rajib Rana (rajibrana.org)

A noise map facilitates monitoring of environmental noise pollution in urban areas. It can raise ... more A noise map facilitates monitoring of environmental noise pollution in urban areas. It can raise citizen awareness of noise pollution levels, and aid in the development of mitigation strategies to cope with the adverse effects. However, state-of-the-art techniques for rendering noise maps in urban areas are expensive and rarely updated (months or even years), as they rely on population and traffic models rather than on real data. Participatory urban sensing can be leveraged to create an open and inexpensive platform for rendering up-to-date noise maps. In this paper, we present the design, implementation and performance eval- uation of an end-to-end participatory urban noise mapping system called Ear- Phone. Ear-Phone, for the first time, leverages Compressive Sensing to ad- dress the fundamental problem of recovering the noise map from incomplete and random samples obtained by crowdsourcing data collection. Ear-Phone, implemented on Nokia N95 and HP iPAQ mobile devices, also addres...

Real-time classification via sparse representation in acoustic sensor networks

by Mingrui Yang, Rajib Rana (rajibrana.org), and Chun Chou

Proceedings of the 11th ACM Conference on Embedded Networked Sensor Systems - SenSys '13, 2013

Acoustic Sensor Networks (ASNs) have a wide range of applications in natural and urban environmen... more Acoustic Sensor Networks (ASNs) have a wide range of applications in natural and urban environment monitoring, as well as indoor activity monitoring. In-network classification is critically important in ASNs because wireless transmission costs several orders of magnitude more energy than computation. The main challenges of in-network classification in ASNs include effective feature selection, intensive computation requirement and high noise levels. To address these challenges, we propose a sparse representation based featureless, low computational cost, and noise resilient framework for in-network classification in ASNs. The key component of Sparse Approximation based Classification (SAC), 1 minimization, is a convex optimization problem, and is known to be computationally expensive. Furthermore, SAC algorithms assumes that the test samples are a linear combination of a few training samples in the training sets. For acoustic applications, this results in a very large training dictionary, making the computation infeasible to be performed on resource constrained ASN platforms. Therefore, we propose several techniques to reduce the size of the problem, so as to fit SAC for in-network classification in ASNs. Our extensive evaluation using two real-life datasets (consisting of calls from 14 frog species and 20 cricket species respectively) shows that the proposed SAC framework outperforms conventional approaches such as Support Vector Machines (SVMs) and k-Nearest Neighbor (kNN) in terms of classification accuracy and robustness. Moreover, our SAC approach can deal with multi-label classification which is common in ASNs. Finally, we explore the system design spaces and demonstrate the real-time feasibility of the proposed framework by the implementation and evaluation of an acoustic classification application on an embedded ASN testbed.

Distributed sparse approximation for frog sound classification

by Mingrui Yang and Rajib Rana (rajibrana.org)

Proceedings of the 11th international conference on Information Processing in Sensor Networks - IPSN '12, 2012

Sparse approximation has now become a buzzword for classification in numerous research domains. W... more Sparse approximation has now become a buzzword for classification in numerous research domains. We propose a distributed sparse approximation method based on 1 minimization for frog sound classification, which is tailored to the resource constrained wireless sensor networks. Our pilot study demonstrates that 1 minimization can run on wireless sensor nodes producing satisfactory classification accuracy.

Reconstruction of temporal-spatial profile from participatory sensing data

The reconstruction of an unknown temporal-spatial profile from participatory sensing data poses a... more The reconstruction of an unknown temporal-spatial profile from participatory sensing data poses a number of challenges due to uncoordinated user movement and possibly low user involvement. This paper considers the problem of reconstructing such a profile from participatory sensing data by exploiting the theory of compressive sensing. In particular we study the impact of the number of users and opportunistic user communication on the reconstruction accuracy. Our simulation results show that, in the case where each user is limited to return one projection, a small amount of local data exchange between the users can improve the reconstruction accuracy.

An adaptive algorithm for compressive approximation of trajectory (aacat) for delay tolerant networks

Highly efficient compression provides a promising approach to address the transmission and comput... more Highly efficient compression provides a promising approach to address the transmission and computation challenges imposed by moving object tracking applications on resource constrained Wireless Sensor Networks (WSNs). In this paper, we propose and design a Compressive Sensing (CS) based trajectory approximation algorithm, Adaptive Algorithm for Compressive Approximation of Trajectory (AACAT), which performs trajectory compression, so as to maximize the information about the trajectory subject to limited bandwidth. Our extensive evaluation using “real” trajectories of three different object groups (animals, pedestrians and vehicles) shows that CS-based trajectory compression reduces up to 30% transmission overheads, for given information loss bounds, compared to the state-of-the-art trajectory compression algorithms. We implement AACAT on the resource-impoverished sensor nodes, which shows that AACAT achieves high compression performance with very limited resource (computation power and energy) overheads.

ENERGY-AWARE COMPRESSIVE SENSING

Compressive sensing for gait recognition

Compressive Sensing (CS) is a popular signal processing technique, that can exactly reconstruct a... more Compressive Sensing (CS) is a popular signal processing technique, that can exactly reconstruct a signal given a small number of random projections of the original signal, provided that the signal is sufficiently sparse. We demonstrate the applicability of CS in the field of gait recognition as a very effective dimensionality reduction technique, using the gait energy image (GEI) as the feature extraction process. We compare the CS based approach to the principal component analysis (PCA) and show that the proposed method outperforms this baseline, particularly under situations where there are appearance changes in the subject. Applying CS to the gait features also avoids the need to train the models, by using a generalised random projection.

EarPhone assessment of noise pollution with mobile phones

EEG source localization using a sparsity prior based on Brodmann areas

by Rajib Rana (rajibrana.org), Frank De Hoog, and S. Saha

Localizing the sources of electrical activity in the brain from Electroencephalographic (EEG) dat... more Localizing the sources of electrical activity in the brain from Electroencephalographic (EEG) data is an important tool for non-invasive study of brain dynamics. Generally, the source localization process involves a high-dimensional inverse problem that has an infinite number of solutions and thus requires additional constraints to be considered to have a unique solution. In the context of EEG source localization, we propose a novel approach that is based on dividing the cerebral cortex of the brain into a finite number of "Functional Zones" which correspond to unitary functional areas in the brain. In this paper we investigate the use of Brodmann's areas as the Functional Zones. This approach allows us to apply a sparsity constraint to find a unique solution for the inverse EEG problem. Compared to previously published algorithms which use different sparsity constraints to solve this problem, the proposed method is potentially more consistent with the known sparsity profile of the human brain activity and thus may be able to ensure better localization. Numerical experiments are conducted on a realistic head model obtained from segmentation of MRI images of the head and includes four major compartments namely scalp, skull, cerebrospinal fluid (CSF) and brain with relative conductivity values. Three different electrode setups are tested in the numerical experiments.

Sparse temporal representations for facial expression recognition

In automatic facial expression recognition, an increasing number of techniques had been proposed ... more In automatic facial expression recognition, an increasing number of techniques had been proposed for in the literature that exploits the temporal nature of facial expressions. As all facial expressions are known to evolve over time, it is crucially important for a classifier to be capable of modelling their dynamics. We establish that the method of sparse representation (SR) classifiers proves to be a suitable candidate for this purpose, and subsequently propose a framework for expression dynamics to be efficiently incorporated into its current formulation. We additionally show that for the SR method to be applied effectively, then a certain threshold on image dimensionality must be enforced (unlike in facial recognition problems). Thirdly, we determined that recognition rates may be significantly influenced by the size of the projection matrix Φ. To demonstrate these, a battery of experiments had been conducted on the CK+ dataset for the recognition of the seven prototypic expressions − anger, contempt, disgust, fear, happiness, sadness and surprise − and comparisons have been made between the proposed temporal-SR against the static-SR framework and state-of-the-art support vector machine.

Dynamic texture reconstruction from sparse codes for unusual event detection in crowded scenes

Unusual event detection in crowded scenes remains challenging because of the diversity of events ... more Unusual event detection in crowded scenes remains challenging because of the diversity of events and noise. In this paper, we present a novel approach for unusual event detection via sparse reconstruction of dynamic textures over an overcomplete basis set, with the dynamic texture described by local binary patterns from three orthogonal planes (LBP-TOP). The overcomplete basis set is learnt from the training data where only the normal items observed. In the detection process, given a new observation, we compute the sparse coefficients using the Dantzig Selector algorithm which was proposed in the literature of compressed sensing. Then the reconstruction errors are computed, based on which we detect the abnormal items. Our application can be used to detect both local and global abnormal events. We evaluate our algorithm on UCSD Abnormality Datasets for local anomaly detection, which is shown to outperform current state-ofthe-art approaches, and we also get promising results for rapid escape detection using the PETS2009 dataset.

Passive In-home Estimation of Gait Velocity From Motion Sensors

Gait velocity has been consistently shown to be an important indicator and predictor of health st... more Gait velocity has been consistently shown to be an important indicator and predictor of health status, especially in elderly populations. Gait velocity is often assessed clinically, but clinical assessments occur infrequently and thus do not allow optimal detection of key health changes when they occur. In this paper, we propose using the time it takes a person to move between rooms in their home -referred to as "transition times" -estimated from passive infrared motion sensors installed in a patients own home to predict gait velocity. By using a support vector regression approach to model the relationship between transition times and gait velocities, we show that we can predict velocities accurately. In particular, we demonstrate that the proposed approach has an average error of 4 cm/sec using data collected over a 5 year period from 76 study participants monitored both in their own homes and within a clinical setting. The proposed method is simple and cost effective, and has advantages over competing approaches such as not requiring patients to wear a device or needing dedicated sensors to measure gait. Furthermore, this method provides substantially more frequent estimates of gait velocity than are provided by other approaches.

EarPhone: A context-aware noise mapping using smart phones

by Rajib Rana (rajibrana.org) and Nirupama Bulusu

A noise map facilitates the monitoring of environmental noise pollution in urban areas. It can ra... more A noise map facilitates the monitoring of environmental noise pollution in urban areas. It can raise citizen awareness of noise pollution levels, and aid in the development of mitigation strategies to cope with the adverse effects. However, state-of-the-art techniques for rendering noise maps in urban areas are expensive and rarely updated (for months or even years), as they rely on population and traffic models rather than on real data. Smart phone based urban sensing can be leveraged to create an open and inexpensive platform for rendering up-todate noise maps. In this paper, we present the design, implementation and performance evaluation of an end-to-end, context-aware, noise mapping system called Ear-Phone.

Signal Reconstruction from Rechargeable Wireless Sensor Networks using Sparse Random Projections

Due to non-homogeneous spread of sunlight, sensing nodes possess non-uniform energy budget in rec... more Due to non-homogeneous spread of sunlight, sensing nodes possess non-uniform energy budget in rechargeable Wireless Sensor Networks (WSNs). An energy-aware workload distribution strategy is therefore necessary to achieve good data accuracy subject to energy-neutral operation. Our previously proposed Energy Aware Sparse approximation Technique (EAST) can approximate a signal, by adapting sensor node sampling workload according to solar energy availability. However, the major shortcoming of EAST is that it does not guarantee an optimal sensing strategy. In other words EAST offers energy neutral operation, however it does not offer the best utilization of sensor node energy, which compromises the reconstruction accuracy. In order to overcome this shortcoming, we propose EAST + which, maximizes the reconstruction accuracy subject to energy neutral operations. We also propose a distributed algorithm for EAST + , which offers accurate signal reconstruction with limited node to-base communications.

Cross-layer interactions in energy efficient information collection in wireless sensor networks with adaptive compressive sensing

We consider the problem of using Wireless Sensor Networks (WSNs) to measure the temporal-spatial ... more We consider the problem of using Wireless Sensor Networks (WSNs) to measure the temporal-spatial field of some scalar physical quantities. Our goal is to obtain a sufficiently accurate approximation of the temporal-spatial field with as little energy as possible. We propose an adaptive algorithm, based on the recently developed theory of adaptive compressive sensing, to collect information from WSNs in an energy efficient manner. The key idea of the algorithm is to perform "projections" iteratively to maximise the amount of information gain per energy expenditure. We prove that this maximisation problem is NP-hard and propose a number of heuristics to solve this problem. This maximisation problem can also be viewed as a routing problem with a metric which is a non-linear function of the data field; therefore the problem is cross-layer in nature, requiring information in both application and network layers. We evaluate the performance of our proposed algorithms using data from both simulation and an outdoor WSN testbed. The results show that our proposed algorithms are able to give a more accurate approximation of the temporal-spatial field for a given energy expenditure.

Real-time classification via sparse representation in acoustic sensor networks

by Rajib Rana (rajibrana.org), Mingrui Yang, and B. Wei

Acoustic Sensor Networks (ASNs) have a wide range of applications in natural and urban environmen... more Acoustic Sensor Networks (ASNs) have a wide range of applications in natural and urban environment monitoring, as well as indoor activity monitoring. In-network classification is critically important in ASNs because wireless transmission costs several orders of magnitude more energy than computation. The main challenges of in-network classification in ASNs include effective feature selection, intensive computation requirement and high noise levels. To address these challenges, we propose a sparse representation based featureless, low computational cost, and noise resilient framework for in-network classification in ASNs. The key component of Sparse Approximation based Classification (SAC), 1 minimization, is a convex optimization problem, and is known to be computationally expensive. Furthermore, SAC algorithms assumes that the test samples are a linear combination of a few training samples in the training sets. For acoustic applications, this results in a very large training dictionary, making the computation infeasible to be performed on resource constrained ASN platforms. Therefore, we propose several techniques to reduce the size of the problem, so as to fit SAC for in-network classification in ASNs. Our extensive evaluation using two real-life datasets (consisting of calls from 14 frog species and 20 cricket species respectively) shows that the proposed SAC framework outperforms conventional approaches such as Support Vector Machines (SVMs) and k-Nearest Neighbor (kNN) in terms of classification accuracy and robustness. Moreover, our SAC approach can deal with multi-label classification which is common in ASNs. Finally, we explore the system design spaces and demonstrate the real-time feasibility of the proposed framework by the implementation and evaluation of an acoustic classification application on an embedded ASN testbed.

Non-uniform compressive sensing in wireless sensor networks: Feasibility and application

In this paper, we consider the problem of using wireless sensor networks (WSNs) to measure the te... more In this paper, we consider the problem of using wireless sensor networks (WSNs) to measure the temporal-spatial profile of some physical phenomena. We base our work on two observations. Firstly, most physical phenomena are compressible in some transform domain basis. Secondly, most WSNs have some form of heterogeneity. Given these two observations, we propose a non-uniform compressive sensing method to improve the performance of WSNs by exploiting both compressibility and heterogeneity. We apply our proposed method to a real WSN data set. We find that our method can provide a more accurate temporal-spatial profile for a given energy budget compared with other sampling methods.

by Nirupama Bulusu, Chun Chou, and Rajib Rana (rajibrana.org)

Proceedings of the 9th ACM/IEEE International Conference on Information Processing in Sensor Networks - IPSN '10, 2010

A noise map facilitates monitoring of environmental noise pollution in urban areas. It can raise ... more A noise map facilitates monitoring of environmental noise pollution in urban areas. It can raise citizen awareness of noise pollution levels, and aid in the development of mitigation strategies to cope with the adverse effects. However, state-of-the-art techniques for rendering noise maps in urban areas are expensive and rarely updated (months or even years), as they rely on population and traffic models rather than on real data. Participatory urban sensing can be leveraged to create an open and inexpensive platform for rendering up-to-date noise maps.

Nonuniform compressive sensing for heterogeneous wireless sensor networks

by Chun Chou and Rajib Rana (rajibrana.org)

In this paper, we consider the problem of using wireless sensor networks (WSNs) to measure the te... more In this paper, we consider the problem of using wireless sensor networks (WSNs) to measure the temporal-spatial profile of some physical phenomena. We base our work on two observations. First, most physical phenomena are compressible in some transform domain basis. Second, most WSNs have some form of heterogeneity. Given these two observations, we propose a nonuniform compressive sensing method to improve the performance of WSNs by exploiting both compressibility and heterogeneity. We apply our proposed method to real WSN data sets. We find that our method can provide a more accurate temporal-spatial profile for a given energy budget compared with other sampling methods.

Non-uniform compressive sensing in wireless sensor networks: Feasibility and application

by Chun Chou, Rajib Rana (rajibrana.org), and W. Hu

In this paper, we consider the problem of using wireless sensor networks (WSNs) to measure the te... more In this paper, we consider the problem of using wireless sensor networks (WSNs) to measure the temporal-spatial profile of some physical phenomena. We base our work on two observations. Firstly, most physical phenomena are compressible in some transform domain basis. Secondly, most WSNs have some form of heterogeneity. Given these two observations, we propose a non-uniform compressive sensing method to improve the performance of WSNs by exploiting both compressibility and heterogeneity. We apply our proposed method to a real WSN data set. We find that our method can provide a more accurate temporal-spatial profile for a given energy budget compared with other sampling methods.

Ear-Phone: An End-to-End Participatory Urban Noise Mapping System

by Chun Chou, Nirupama Bulusu, and Rajib Rana (rajibrana.org)

A noise map facilitates monitoring of environmental noise pollution in urban areas. It can raise ... more A noise map facilitates monitoring of environmental noise pollution in urban areas. It can raise citizen awareness of noise pollution levels, and aid in the development of mitigation strategies to cope with the adverse effects. However, state-of-the-art techniques for rendering noise maps in urban areas are expensive and rarely updated (months or even years), as they rely on population and traffic models rather than on real data. Participatory urban sensing can be leveraged to create an open and inexpensive platform for rendering up-to-date noise maps. In this paper, we present the design, implementation and performance eval- uation of an end-to-end participatory urban noise mapping system called Ear- Phone. Ear-Phone, for the first time, leverages Compressive Sensing to ad- dress the fundamental problem of recovering the noise map from incomplete and random samples obtained by crowdsourcing data collection. Ear-Phone, implemented on Nokia N95 and HP iPAQ mobile devices, also addres...

Real-time classification via sparse representation in acoustic sensor networks

by Mingrui Yang, Rajib Rana (rajibrana.org), and Chun Chou

Proceedings of the 11th ACM Conference on Embedded Networked Sensor Systems - SenSys '13, 2013

Acoustic Sensor Networks (ASNs) have a wide range of applications in natural and urban environmen... more Acoustic Sensor Networks (ASNs) have a wide range of applications in natural and urban environment monitoring, as well as indoor activity monitoring. In-network classification is critically important in ASNs because wireless transmission costs several orders of magnitude more energy than computation. The main challenges of in-network classification in ASNs include effective feature selection, intensive computation requirement and high noise levels. To address these challenges, we propose a sparse representation based featureless, low computational cost, and noise resilient framework for in-network classification in ASNs. The key component of Sparse Approximation based Classification (SAC), 1 minimization, is a convex optimization problem, and is known to be computationally expensive. Furthermore, SAC algorithms assumes that the test samples are a linear combination of a few training samples in the training sets. For acoustic applications, this results in a very large training dictionary, making the computation infeasible to be performed on resource constrained ASN platforms. Therefore, we propose several techniques to reduce the size of the problem, so as to fit SAC for in-network classification in ASNs. Our extensive evaluation using two real-life datasets (consisting of calls from 14 frog species and 20 cricket species respectively) shows that the proposed SAC framework outperforms conventional approaches such as Support Vector Machines (SVMs) and k-Nearest Neighbor (kNN) in terms of classification accuracy and robustness. Moreover, our SAC approach can deal with multi-label classification which is common in ASNs. Finally, we explore the system design spaces and demonstrate the real-time feasibility of the proposed framework by the implementation and evaluation of an acoustic classification application on an embedded ASN testbed.

Distributed sparse approximation for frog sound classification

by Mingrui Yang and Rajib Rana (rajibrana.org)

Proceedings of the 11th international conference on Information Processing in Sensor Networks - IPSN '12, 2012

Sparse approximation has now become a buzzword for classification in numerous research domains. W... more Sparse approximation has now become a buzzword for classification in numerous research domains. We propose a distributed sparse approximation method based on 1 minimization for frog sound classification, which is tailored to the resource constrained wireless sensor networks. Our pilot study demonstrates that 1 minimization can run on wireless sensor nodes producing satisfactory classification accuracy.

Reconstruction of temporal-spatial profile from participatory sensing data

The reconstruction of an unknown temporal-spatial profile from participatory sensing data poses a... more The reconstruction of an unknown temporal-spatial profile from participatory sensing data poses a number of challenges due to uncoordinated user movement and possibly low user involvement. This paper considers the problem of reconstructing such a profile from participatory sensing data by exploiting the theory of compressive sensing. In particular we study the impact of the number of users and opportunistic user communication on the reconstruction accuracy. Our simulation results show that, in the case where each user is limited to return one projection, a small amount of local data exchange between the users can improve the reconstruction accuracy.

An adaptive algorithm for compressive approximation of trajectory (aacat) for delay tolerant networks

Highly efficient compression provides a promising approach to address the transmission and comput... more Highly efficient compression provides a promising approach to address the transmission and computation challenges imposed by moving object tracking applications on resource constrained Wireless Sensor Networks (WSNs). In this paper, we propose and design a Compressive Sensing (CS) based trajectory approximation algorithm, Adaptive Algorithm for Compressive Approximation of Trajectory (AACAT), which performs trajectory compression, so as to maximize the information about the trajectory subject to limited bandwidth. Our extensive evaluation using “real” trajectories of three different object groups (animals, pedestrians and vehicles) shows that CS-based trajectory compression reduces up to 30% transmission overheads, for given information loss bounds, compared to the state-of-the-art trajectory compression algorithms. We implement AACAT on the resource-impoverished sensor nodes, which shows that AACAT achieves high compression performance with very limited resource (computation power and energy) overheads.

ENERGY-AWARE COMPRESSIVE SENSING

Compressive sensing for gait recognition

Compressive Sensing (CS) is a popular signal processing technique, that can exactly reconstruct a... more Compressive Sensing (CS) is a popular signal processing technique, that can exactly reconstruct a signal given a small number of random projections of the original signal, provided that the signal is sufficiently sparse. We demonstrate the applicability of CS in the field of gait recognition as a very effective dimensionality reduction technique, using the gait energy image (GEI) as the feature extraction process. We compare the CS based approach to the principal component analysis (PCA) and show that the proposed method outperforms this baseline, particularly under situations where there are appearance changes in the subject. Applying CS to the gait features also avoids the need to train the models, by using a generalised random projection.

EarPhone assessment of noise pollution with mobile phones

EEG source localization using a sparsity prior based on Brodmann areas

by Rajib Rana (rajibrana.org), Frank De Hoog, and S. Saha

Localizing the sources of electrical activity in the brain from Electroencephalographic (EEG) dat... more Localizing the sources of electrical activity in the brain from Electroencephalographic (EEG) data is an important tool for non-invasive study of brain dynamics. Generally, the source localization process involves a high-dimensional inverse problem that has an infinite number of solutions and thus requires additional constraints to be considered to have a unique solution. In the context of EEG source localization, we propose a novel approach that is based on dividing the cerebral cortex of the brain into a finite number of "Functional Zones" which correspond to unitary functional areas in the brain. In this paper we investigate the use of Brodmann's areas as the Functional Zones. This approach allows us to apply a sparsity constraint to find a unique solution for the inverse EEG problem. Compared to previously published algorithms which use different sparsity constraints to solve this problem, the proposed method is potentially more consistent with the known sparsity profile of the human brain activity and thus may be able to ensure better localization. Numerical experiments are conducted on a realistic head model obtained from segmentation of MRI images of the head and includes four major compartments namely scalp, skull, cerebrospinal fluid (CSF) and brain with relative conductivity values. Three different electrode setups are tested in the numerical experiments.

Sparse temporal representations for facial expression recognition

In automatic facial expression recognition, an increasing number of techniques had been proposed ... more In automatic facial expression recognition, an increasing number of techniques had been proposed for in the literature that exploits the temporal nature of facial expressions. As all facial expressions are known to evolve over time, it is crucially important for a classifier to be capable of modelling their dynamics. We establish that the method of sparse representation (SR) classifiers proves to be a suitable candidate for this purpose, and subsequently propose a framework for expression dynamics to be efficiently incorporated into its current formulation. We additionally show that for the SR method to be applied effectively, then a certain threshold on image dimensionality must be enforced (unlike in facial recognition problems). Thirdly, we determined that recognition rates may be significantly influenced by the size of the projection matrix Φ. To demonstrate these, a battery of experiments had been conducted on the CK+ dataset for the recognition of the seven prototypic expressions − anger, contempt, disgust, fear, happiness, sadness and surprise − and comparisons have been made between the proposed temporal-SR against the static-SR framework and state-of-the-art support vector machine.

Dynamic texture reconstruction from sparse codes for unusual event detection in crowded scenes

Unusual event detection in crowded scenes remains challenging because of the diversity of events ... more Unusual event detection in crowded scenes remains challenging because of the diversity of events and noise. In this paper, we present a novel approach for unusual event detection via sparse reconstruction of dynamic textures over an overcomplete basis set, with the dynamic texture described by local binary patterns from three orthogonal planes (LBP-TOP). The overcomplete basis set is learnt from the training data where only the normal items observed. In the detection process, given a new observation, we compute the sparse coefficients using the Dantzig Selector algorithm which was proposed in the literature of compressed sensing. Then the reconstruction errors are computed, based on which we detect the abnormal items. Our application can be used to detect both local and global abnormal events. We evaluate our algorithm on UCSD Abnormality Datasets for local anomaly detection, which is shown to outperform current state-ofthe-art approaches, and we also get promising results for rapid escape detection using the PETS2009 dataset.

Passive In-home Estimation of Gait Velocity From Motion Sensors

Gait velocity has been consistently shown to be an important indicator and predictor of health st... more Gait velocity has been consistently shown to be an important indicator and predictor of health status, especially in elderly populations. Gait velocity is often assessed clinically, but clinical assessments occur infrequently and thus do not allow optimal detection of key health changes when they occur. In this paper, we propose using the time it takes a person to move between rooms in their home -referred to as "transition times" -estimated from passive infrared motion sensors installed in a patients own home to predict gait velocity. By using a support vector regression approach to model the relationship between transition times and gait velocities, we show that we can predict velocities accurately. In particular, we demonstrate that the proposed approach has an average error of 4 cm/sec using data collected over a 5 year period from 76 study participants monitored both in their own homes and within a clinical setting. The proposed method is simple and cost effective, and has advantages over competing approaches such as not requiring patients to wear a device or needing dedicated sensors to measure gait. Furthermore, this method provides substantially more frequent estimates of gait velocity than are provided by other approaches.

EarPhone: A context-aware noise mapping using smart phones

by Rajib Rana (rajibrana.org) and Nirupama Bulusu

A noise map facilitates the monitoring of environmental noise pollution in urban areas. It can ra... more A noise map facilitates the monitoring of environmental noise pollution in urban areas. It can raise citizen awareness of noise pollution levels, and aid in the development of mitigation strategies to cope with the adverse effects. However, state-of-the-art techniques for rendering noise maps in urban areas are expensive and rarely updated (for months or even years), as they rely on population and traffic models rather than on real data. Smart phone based urban sensing can be leveraged to create an open and inexpensive platform for rendering up-todate noise maps. In this paper, we present the design, implementation and performance evaluation of an end-to-end, context-aware, noise mapping system called Ear-Phone.

Signal Reconstruction from Rechargeable Wireless Sensor Networks using Sparse Random Projections

Due to non-homogeneous spread of sunlight, sensing nodes possess non-uniform energy budget in rec... more Due to non-homogeneous spread of sunlight, sensing nodes possess non-uniform energy budget in rechargeable Wireless Sensor Networks (WSNs). An energy-aware workload distribution strategy is therefore necessary to achieve good data accuracy subject to energy-neutral operation. Our previously proposed Energy Aware Sparse approximation Technique (EAST) can approximate a signal, by adapting sensor node sampling workload according to solar energy availability. However, the major shortcoming of EAST is that it does not guarantee an optimal sensing strategy. In other words EAST offers energy neutral operation, however it does not offer the best utilization of sensor node energy, which compromises the reconstruction accuracy. In order to overcome this shortcoming, we propose EAST + which, maximizes the reconstruction accuracy subject to energy neutral operations. We also propose a distributed algorithm for EAST + , which offers accurate signal reconstruction with limited node to-base communications.

Cross-layer interactions in energy efficient information collection in wireless sensor networks with adaptive compressive sensing

We consider the problem of using Wireless Sensor Networks (WSNs) to measure the temporal-spatial ... more We consider the problem of using Wireless Sensor Networks (WSNs) to measure the temporal-spatial field of some scalar physical quantities. Our goal is to obtain a sufficiently accurate approximation of the temporal-spatial field with as little energy as possible. We propose an adaptive algorithm, based on the recently developed theory of adaptive compressive sensing, to collect information from WSNs in an energy efficient manner. The key idea of the algorithm is to perform "projections" iteratively to maximise the amount of information gain per energy expenditure. We prove that this maximisation problem is NP-hard and propose a number of heuristics to solve this problem. This maximisation problem can also be viewed as a routing problem with a metric which is a non-linear function of the data field; therefore the problem is cross-layer in nature, requiring information in both application and network layers. We evaluate the performance of our proposed algorithms using data from both simulation and an outdoor WSN testbed. The results show that our proposed algorithms are able to give a more accurate approximation of the temporal-spatial field for a given energy expenditure.

Real-time classification via sparse representation in acoustic sensor networks

by Rajib Rana (rajibrana.org), Mingrui Yang, and B. Wei

Acoustic Sensor Networks (ASNs) have a wide range of applications in natural and urban environmen... more Acoustic Sensor Networks (ASNs) have a wide range of applications in natural and urban environment monitoring, as well as indoor activity monitoring. In-network classification is critically important in ASNs because wireless transmission costs several orders of magnitude more energy than computation. The main challenges of in-network classification in ASNs include effective feature selection, intensive computation requirement and high noise levels. To address these challenges, we propose a sparse representation based featureless, low computational cost, and noise resilient framework for in-network classification in ASNs. The key component of Sparse Approximation based Classification (SAC), 1 minimization, is a convex optimization problem, and is known to be computationally expensive. Furthermore, SAC algorithms assumes that the test samples are a linear combination of a few training samples in the training sets. For acoustic applications, this results in a very large training dictionary, making the computation infeasible to be performed on resource constrained ASN platforms. Therefore, we propose several techniques to reduce the size of the problem, so as to fit SAC for in-network classification in ASNs. Our extensive evaluation using two real-life datasets (consisting of calls from 14 frog species and 20 cricket species respectively) shows that the proposed SAC framework outperforms conventional approaches such as Support Vector Machines (SVMs) and k-Nearest Neighbor (kNN) in terms of classification accuracy and robustness. Moreover, our SAC approach can deal with multi-label classification which is common in ASNs. Finally, we explore the system design spaces and demonstrate the real-time feasibility of the proposed framework by the implementation and evaluation of an acoustic classification application on an embedded ASN testbed.

Non-uniform compressive sensing in wireless sensor networks: Feasibility and application

In this paper, we consider the problem of using wireless sensor networks (WSNs) to measure the te... more In this paper, we consider the problem of using wireless sensor networks (WSNs) to measure the temporal-spatial profile of some physical phenomena. We base our work on two observations. Firstly, most physical phenomena are compressible in some transform domain basis. Secondly, most WSNs have some form of heterogeneity. Given these two observations, we propose a non-uniform compressive sensing method to improve the performance of WSNs by exploiting both compressibility and heterogeneity. We apply our proposed method to a real WSN data set. We find that our method can provide a more accurate temporal-spatial profile for a given energy budget compared with other sampling methods.