Research Interests


Research Projects and Impacts


Internet of Things (IoT) Research

Internet of Things (IoT) Currently there are about 20 billion IoT devices connected to the Internet. By 2020, an estimated 50 billion devices will be connected. Already 40% of the 7 billion population have internet connection. By 2050, 70% of the world's population and over 6 billion people are expected to live in cities and surrounding regions. Managing city, people and resources (water, electricity, air, land, transport, public health) are set to become challenging. Technical challenges include designing new embedded IoT devices, communication protocols, network management, analytics, security and privacy, scalability, sustainability and many other. Our group researches on these challenging IoT issues in collaboration with academia, industry and governments.

Details and Outcomes

Grants: ARC Linkage Projects (LP120100529)- "Creating Smart City through Internet of Things"; ARC LIEF (LE120100129) - "Internet of Things Testbed for Creating a Smart City"; EU FP7 SocioTal - "Creating a Socially Aware Citizen-centric Internet of Things" and H2020 Organicity - "Co-creating Smart Cities of the Future"

Collaborators: The University of Sydney, Australia; Swinburne University of technology, Australia; Deakin University, Australia; Nanjing University of Posts and Telecommunications, China; University of Surrey, UK; Institute for Infocomm Research, Singapore;

Partners: Defence Science & Technology (DSTO); City of Melbourne; Arup; SenSen Networks, Pty Ltd

Outcomes:IoT framework for creating a smart city, QoS solution using IoT, framework for handling IoT issues, such as security, privacy and governance.

Publications:

  1. J. Gubbi, R. Buyya, S. Marusic, and M. Palaniswami, "Internet of Things (IoT): A vision, architectural elements, and future directions," Future Generation Computer Systems, vol. 29, pp. 1645-1660, 2013. (highly cited paper in IoT)
  2. J. Jin, J. Gubbi, S. Marusic, and M. Palaniswami, "An information framework of creating a smart city through Internet of things," IEEE Internet of Things Journal, vol. 1, pp. 112-121, 2014.
  3. J. Jin, M. Palaniswami, D. Yuan, Y.-N. Dong, and K. Moessner, "Priority service provisioning and max–min fairness: A utility-based flow control approach," Journal of Network and Systems Management, pp. 1-19, 2016.
  4. S. Marusic, J. Gubbi, H. Sullivan, Y. W. Law and, and M. Palaniswami, "Participatory Sensing, Privacy, and Trust Management for Interactive Local Government," IEEE Technology and Society Magazine, vol. 33, pp. 62-70, 2014.
  5. J. Jin, J. Gubbi, T. Luo, and M. Palaniswami, "Network architecture and QoS issues in the Internet of things for a smart city," in The 12th International Symposium on Communications and Information Technologies (ISCIT 2012), 2012.
Event Analytics in Big Data: The current Internet is evolving into the Internet of Things (IoT), where networked devices have the ability to compute, sense and interact with their surroundings. IoT deployments will generate a vast amount of data, but the value of the data lies in how we can exploit it. Currently there is little domain expertise to automate this Big data analysis, and traditional supervised machine learning techniques suffer from a lack of labelled training data in this context. The aim is to develop scalable distributed and incremental event detection techniques in addition to privacy-preserving data analytics solutions for the IoT. Also enables a new generation of applications in IoT applications such as smart cities. We have developed algorithms to detect anomalies in traffic patterns of car using taxi data. Novel incremental learning algorithms have been designed and developed to this end. The solution can be used to monitor parks, urban canopies, plants and tree species. We have also developed a solution with state-of-the-art interactive visualization for new touch-enables devices such as smartphones and tablets. This provides councils, administrators and policymakers to assess the situation in real-time, enabling to make informed and immediate decisions.

Details and Outcomes

Grants: ARC Linkage Projects (LP120100529)- "Creating Smart City through Internet of Things"; ARC LIEF (LE120100129) - "Internet of Things Testbed for Creating a Smart City"; EU FP7 SocioTal - "Creating a Socially Aware Citizen-centric Internet of Things" and H2020 Organicity - "Co-creating Smart Cities of the Future"

Collaborators: University of Surrey, UK; Michigan Technological University, USA; Deakin University, Australia;

Partners: Defence Science & Technology (DSTO); City of Melbourne; Arup; SenSen Networks, Pty Ltd

Outcomes: Real-time event analytics capable of detecting anomalies from IoT data, sensor drift correction, incrementally learning from streaming data, spatiotemporal estimation of missing data, capability to interact with users for real-time decision making.

  1. Y. Zheng, S. Rajasegarar, and C. Leckie, "Parking availability prediction for sensor-enabled car parks in smart cities," in Intelligent Sensors, Sensor Networks and Information Processing (ISSNIP), 2015 IEEE Tenth International Conference on, 2015, pp. 1-6.
  2. A. Shilton, S. Rajasegarar, C. Leckie, and M. Palaniswami, "DP1SVM: A dynamic planar one-class support vector machine for Internet of Things environment," in Recent Advances in Internet of Things (RIoT), 2015 International Conference on, 2015, pp. 1-6.
  3. D. Kumar, S. Rajasegarar, and M. Palaniswami, "Geospatial Estimation-Based Auto Drift Correction in Wireless Sensor Networks," ACM Transactions on Sensor Networks (TOSN), vol. 11, p. 50, 2015.
  4. D. Kumar, J. C. Bezdek, S. Rajasegarar, C. Leckie, and M. Palaniswami, "A visual-numeric approach to clustering and anomaly detection for trajectory data," The Visual Computer, pp. 1-17, 2015.
  5. D. Kumar, J. C. Bezdek, M. Palaniswami, S. Rajasegarar, C. Leckie, and T. C. Havens, "A Hybrid Approach to Clustering in Big Data," IEEE Transactions on Cybernetics, 2015.
  6. Y. Zheng, S. Rajasegarar, C. Leckie, and M. Palaniswami, "Smart car parking: temporal clustering and anomaly detection in urban car parking," in Intelligent Sensors, Sensor Networks and Information Processing (ISSNIP), 2014 IEEE Ninth International Conference on, 2014, pp. 1-6.
  7. L. Rashidi, S. Rajasegarar, C. Leckie, M. Nati, A. Gluhak, M. A. Imran, et al., "Profiling spatial and temporal behaviour in sensor networks: A case study in energy monitoring," in Intelligent Sensors, Sensor Networks and Information Processing (ISSNIP), 2014 IEEE Ninth International Conference on, 2014, pp. 1-7.
  8. S. Rajasegarar, P. Zhang, Y. Zhou, S. Karunasekera, C. Leckie, and M. Palaniswami, "High resolution spatio-temporal monitoring of air pollutants using wireless sensor networks," in Intelligent Sensors, Sensor Networks and Information Processing (ISSNIP), 2014 IEEE Ninth International Conference on, 2014, pp. 1-6.
  9. S. Rajasegarar, C. Leckie, and M. Palaniswami, "Hyperspherical cluster based distributed anomaly detection in wireless sensor networks," Journal of Parallel and Distributed Computing, vol. 74, pp. 1833-1847, 2014.
  10. S. Rajasegarar, C. Leckie, and M. Palaniswami, "Spatio-temporal estimation with Bayesian maximum entropy and compressive sensing in communication constrained networks," in 2014 IEEE International Conference on Communications (ICC), 2014, pp. 4536-4541.
  11. S. Rajasegarar, T. C. Havens, S. Karunasekera, C. Leckie, J. C. Bezdek, M. Jamriska, et al., "High-resolution monitoring of atmospheric pollutants using a system of low-cost sensors," IEEE Transactions on Geoscience and Remote Sensing, vol. 52, pp. 3823-3832, 2014.
  12. S. Rajasegarar, A. Gluhak, M. A. Imran, M. Nati, M. Moshtaghi, C. Leckie, et al., "Ellipsoidal neighbourhood outlier factor for distributed anomaly detection in resource constrained networks," Pattern Recognition, vol. 47, pp. 2867-2879, 2014.
  13. S. M. Erfani, Y. W. Law, S. Karunasekera, C. A. Leckie, and M. Palaniswami, "Privacy-preserving collaborative anomaly detection for participatory sensing," in Pacific-Asia Conference on Knowledge Discovery and Data Mining, 2014, pp. 581-593.
  14. M. Mattess, R. N. Calheiros, and R. Buyya, "Scaling mapreduce applications across hybrid clouds to meet soft deadlines," in Advanced Information Networking and Applications (AINA), 2013 IEEE 27th International Conference on, 2013, pp. 629-636.
  15. D. Kumar, S. Rajasegarar, and M. Palaniswami, "Automatic Sensor drift detection and correction using Spatial Kriging and Kalman filtering," in 2013 IEEE International Conference on Distributed Computing in Sensor Systems, 2013, pp. 183-190.
  16. D. Kumar, M. Palaniswami, S. Rajasegarar, C. Leckie, J. C. Bezdek, and T. C. Havens, "clusiVAT: A mixed visual/numerical clustering algorithm for big data," in Big Data, 2013 IEEE International Conference on, 2013, pp. 112-117.
Energy Monitoring: It is expected that the world energy consumption in residential sector is set to increase by 48% from 2012 to 2040. By 2040, around 13% of world energy consumption would account for homes, buildings and non-transport. We are developing algorithms to monitor energy consumption of the buildings and office spaces to understand the nature of energy consumption. We are also working to develop solutions to intelligently integrate smart grid and residential energy usage.

Details and Outcomes

Collaborators: Deakin University, Australia; Univeristy of Twente, The Netherlands; Delft University of Technology, The Netherlands; Politecnico di Milano, Italy; NICTA Victoria Research Laboratory, Australia; University of Surrey, UK

Outcomes: We have developed optimization techniques and machine learning models for energy monitoring

Publications:

  1. L. Rashidi, S. Rajasegarar, C. Leckie, M. Nati, A. Gluhak, M. A. Imran, et al., "Profiling spatial and temporal behaviour in sensor networks: A case study in energy monitoring Control and communication techniques for the smart grid: An energy efficiency perspective," in IEEE ISSNIP 2014 - 2014 IEEE 9th International Conference on Intelligent Sensors, Sensor Networks and Information Processing, Conference Proceedings, 2014, pp. 987-998.
  2. A. Pirisi, F. Grimaccia, M. Mussetta, R. E. Zich, R. Johnstone, M. Palaniswami, et al., "Optimization of an energy harvesting buoy for coral reef monitoring," in 2013 IEEE Congress on Evolutionary Computation, CEC 2013, 2013, pp. 629-634.
  3. Y. W. Law, T. Alpcan, V. C. S. Lee, A. Lo, S. Marusic, and M. Palaniswami, "Demand response architectures and load management algorithms for energy-efficient power grids: A survey," in Proceedings - 2012 7th International Conference on Knowledge, Information and Creativity Support Systems, KICSS 2012, 2012, pp. 134-141.
  4. A. S. Rao, J. Gubbi, T. Ngo, J. Nguyen, and M. Palaniswami, "Energy efficient time synchronization in WSN for critical infrastructure monitoring," in Communications in Computer and Information Science vol. 197 CCIS, ed, 2011, pp. 314-323.
  5. Y. W. Law, M. Palaniswami, L. V. Hoesel, J. Doumen, P. Hartel, and P. Havinga, "Energy-efficient link-layer jamming attacks against wireless sensor network MAC protocols," ACM Transactions on Sensor Networks, vol. 5, 2009.
  6. Y. W. Law, S. Chatterjea, J. Jin, T. Hanselmann, and M. Palaniswami, "Energy-efficient data acquisition by adaptive sampling for wireless sensor networks," in Proceedings of the 2009 ACM International Wireless Communications and Mobile Computing Conference, IWCMC 2009, 2009, pp. 1146-1151.
  7. S. W. Ekanayake, P. N. Pathirana, B. F. Rolfe, and M. Palaniswami, "Energy efficient, fully-connected mesh networks for high speed applications," in IEEE Vehicular Technology Conference, 2008, pp. 2606-2610.
Security and Privacy in IoT: The IoT consists of islands of sensor networks are meant to operate unattended, potentially in harsh and even hostile environments. Security is thus essential to ensuring the intended operation of the IoT. For businesses and consumers, security enables the IoT to be used with confidence. Privacy is another paradigm where data are collected from citizen engagement and participatory sensing. Clearly, here the data can contain sensitive information, such as participants' location, voice and other proprietary audio-visual information. We are investigating privacy-preserving data mining, that is, we are addressing the important topic of preserving users' privacy to allow a third-party data miner mines user-contributed data for useful information.

Details and Outcomes

Grants: ARC LIEF (LE120100129) - "Internet of Things Testbed for Creating a Smart City"; EU FP7 SmartSantander; EU FP7 SocioTal - "Creating a Socially Aware Citizen-centric Internet of Things" and H2020 Organicity - "Co-creating Smart Cities of the Future"

Collaborators: University of Twente, The Netherlands; National University of Kaohsiung, Taiwan; Università di Roma, Italy; Northwestern Polytechnical University, China; South China Normal University, China; University of Oxford, UK; Delft University of Technology, The Netherlands; University of South Australia, Australia; The University of New South Wales, Australia; Swinburne University of Technology, Australia;

Partners: City of Melbourne; Arup; Agency for Science, Technology and Research, Singapore.

Outcomes: Algorithms, learning models and techniques for security and privacy in critical infrastructure and IoT. These include multicast authentication, secure reprogramming, and resilience to smart jamming, cryptographic key management, secure routing, and secure data aggregation

Publications:

  1. L. Lyu, Y. W. Law, S. M. Erfani, C. Leckie, M. Palaniswami, Y. W. Law, et al., "An improved scheme for privacy-preserving collaborative anomaly detection Security games for risk minimization in automatic generation control," 2016 IEEE International Conference on Pervasive Computing and Communication Workshops, PerCom Workshops 2016, vol. 30, pp. 223-232, 2016.
  2. S. Marusic, J. Gubbi, H. Sullivan, Y. W. Law and, and M. Palaniswami, "Participatory Sensing, Privacy, and Trust Management for Interactive Local Government," IEEE Technology and Society Magazine, vol. 33, pp. 62-70, 2014.
  3. Y. W. Law, H. R. Pota, J. Jin, Z. Man, and M. Palaniswami, "Control and communication techniques for the smart grid: An energy efficiency perspective," in IFAC Proceedings Volumes (IFAC-PapersOnline), 2014, pp. 987-998.
  4. S. M. Erfani, Y. W. Law, S. Karunasekera, C. A. Leckie, and M. Palaniswami, "Privacy-preserving collaborative anomaly detection for participatory sensing," in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) vol. 8443 LNAI, ed, 2014, pp. 581-593.
  5. J. W. Li, S. N. Li, Y. Zhang, Y. W. Law, X. Zhou, and M. Palaniswami, "Analytical model of coding-based reprogramming protocols in lossy wireless sensor networks," in IEEE International Conference on Communications, 2013, pp. 1867-1871.
  6. Y. W. Law, M. Palaniswami, G. Kounga, and A. Lo, "WAKE: Key management scheme for wide-area measurement systems in smart grid," IEEE Communications Magazine, vol. 51, pp. 34-41, 2013.
  7. Y. W. Law, Z. Gong, T. Luo, S. Marusic, and M. Palaniswami, "Comparative study of multicast authentication schemes with application to wide-area measurement system," in ASIA CCS 2013 - Proceedings of the 8th ACM SIGSAC Symposium on Information, Computer and Communications Security, 2013, pp. 287-298.
  8. J. Gubbi, S. Marusic, A. S. Rao, Y. W. Law, and M. Palaniswami, "A pilot study of urban noise monitoring architecture using wireless sensor networks," in Advances in Computing, Communications and Informatics (ICACCI), 2013 International Conference on, 2013, pp. 1047-1052.
  9. L. H. Yen, Y. W. Law, and M. Palaniswami, "Risk-aware distributed beacon scheduling for tree-based ZigBee wireless networks," IEEE Transactions on Mobile Computing, vol. 11, pp. 692-703, 2012.
  10. Y. W. Law, T. Alpcan, M. Palaniswami, and S. Dey, "Security games and risk minimization for automatic generation control in smart grid," in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) vol. 7638 LNCS, ed, 2012, pp. 281-295.
  11. Y. W. Law, T. Alpcan, and M. Palaniswami, "Security games for voltage control in smart grid," in 2012 50th Annual Allerton Conference on Communication, Control, and Computing, Allerton 2012, 2012, pp. 212-219.
  12. Y. Zhang, X. Zhou, Y. W. Law, and M. Palaniswami, "A pollution-resistant method for reprogramming (PRMR) in wireless sensor networks," Xibei Gongye Daxue Xuebao/Journal of Northwestern Polytechnical University, vol. 29, pp. 443-448, 2011.
  13. Y. W. Law, Y. Zhang, J. Jin, M. Palaniswami, and P. Havinga, "Secure rateless deluge: Pollution-resistant reprogramming and data dissemination for wireless sensor networks," Eurasip Journal on Wireless Communications and Networking, vol. 2011, 2011.
  14. Y. W. Law, G. Moniava, Z. Gong, P. Hartel, and M. Palaniswami, "KALwEN: A new practical and interoperable key management scheme for body sensor networks," Security and Communication Networks, vol. 4, pp. 1309-1329, 2011.
  15. Y. Zhang, X. S. Zhou, Y. W. Law, and M. Palaniswami, "Insider DoS attacks on epidemic propagation strategies of network reprogramming in wireless sensor networks," in 5th International Conference on Information Assurance and Security, IAS 2009, 2009, pp. 263-266.
  16. Y. W. Law, L. H. Yen, R. Di Pietro, and M. Palaniswami, "Secure k-connectivity properties of wireless sensor networks," in From Problem toward Solution: Wireless Sensor Networks Security, ed, 2009, pp. 257-270.
  17. Y. W. Law, M. Palaniswami, L. V. Hoesel, J. Doumen, P. Hartel, and P. Havinga, "Energy-efficient link-layer jamming attacks against wireless sensor network MAC protocols," ACM Transactions on Sensor Networks, vol. 5, 2009.
  18. Y. W. Law, S. Chatterjea, J. Jin, T. Hanselmann, and M. Palaniswami, "Energy-efficient data acquisition by adaptive sampling for wireless sensor networks," in Proceedings of the 2009 ACM International Wireless Communications and Mobile Computing Conference, IWCMC 2009, 2009, pp. 1146-1151.

Computer Vision, Image and Video Processing Research

Real-time Video Analytics: The prevalence of camera networks for surveillance, together with the decreasing cost of infrastructure, has produced a significant demand for robust monitoring systems. Current systems offer limited functionality, particularly in their reliance on centralised processing of gathered information. This project addresses end-to-end system challenges of camera networks. Integrating developments across the spatial, spatiotemporal and decision domains, the project incorporates distributed sensor network technology with intelligent formation fusion to deliver unique long-term behaviour analysis in highly crowded environments. We have developed video analytics with capabilities to count people, track, and detect suspicious behaviour, suitable for crowd management, modelling, and urban planning.

Details and Outcomes

Grants: ARC Linkage Projects (LP100200430) - "Design of Adaptive Learning Visual Sensor Networks for Crowd Modelling in High-Density and Occluded Scenarios"

Partners: City of Melbourne, Melbourne Cricket Club, Arup, and SenSen Networks Pty Ltd

Outcomes: Real-time video analytics researched and developed specifically for crowd monitoring in crowded scenes using low-quality videos. It is able to count people, track crowd movements, detect suspicious behaviour (such as loitering). It uses state-of-the art techniques and cutting-edge, novel analytics, unique to crowd monitoring problems.

Publications:

  1. M. Yang, S. Rajasegarar, A. S. Rao, C. Leckie, and M. Palaniswami, "Anomalous Behavior Detection in Crowded Scenes Using Clustering and Spatio-Temporal Features," in Intelligent Information Processing VIII: 9th IFIP TC 12 International Conference, IIP 2016, pp. 132-141.
  2. A. S. Rao, J. Gubbi, and M. Palaniswami, "Anomalous crowd event analysis using isometric mapping," in Advances in Intelligent Systems and Computing vol. 425, ed, 2016, pp. 407-418.
  3. A. S. Rao, J. Gubbi, and M. Palaniswami, "An improved approach to crowd event detection by reducing data dimensions," in Advances in Intelligent Systems and Computing vol. 425, ed, 2016, pp. 85-96.
  4. A. S. Rao, J. Gubbi, S. Marusic, and M. Palaniswami, "Crowd Event Detection on Optical Flow Manifolds," IEEE Transactions on Cybernetics, vol. 46, pp. 1524-1537, 2016.
  5. A. S. Rao, J. Gubbi, S. Rajasegarar, S. Marusic, and M. Palaniswami, "Detection of anomalous crowd behaviour using hyperspherical clustering," in 2014 International Conference on Digital Image Computing: Techniques and Applications, DICTA 2014, 2015.
  6. A. S. Rao, J. Gubbi, S. Marusic, and M. Palaniswami, "Probabilistic detection of crowd events on riemannian manifolds," in 2014 International Conference on Digital Image Computing: Techniques and Applications, DICTA 2014, 2015.
  7. F.-C. Hsu, J. Gubbi, and M. Palaniswami, "Head detection using motion features and multi level pyramid architecture," Computer Vision and Image Understanding, vol. 137, pp. 38-49, 2015.
  8. F.-C. Hsu, J. Gubbi, and M. Palaniswami, "Learning Efficiently-The Deep CNNs-Tree Network," in Digital Image Computing: Techniques and Applications (DICTA), 2015 International Conference on, 2015, pp. 1-7.
  9. A. S. Rao, J. Gubbi, S. Marusic, and M. Palaniswami, "Estimation of crowd density by clustering motion cues," Visual Computer, vol. 31, pp. 1533-1552, 2014.
  10. A. S. Rao, J. Gubbi, S. Marusic, P. Stanley, and M. Palaniswami, "Crowd density estimation based on optical flow and hierarchical clustering," in Proceedings of the 2013 International Conference on Advances in Computing, Communications and Informatics, ICACCI 2013, 2013, pp. 494-499.
  11. A. S. Rao, J. Gubbi, S. Marusic, A. Maher, and M. Palaniswami, "Determination of object directions using optical flow for crowd monitoring," in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) vol. 8034 LNCS, ed, 2013, pp. 613-622.
  12. F.-C. Hsu, J. Gubbi, and M. Palaniswami, "Human head detection using Histograms of Oriented optical flow in low quality videos with occlusion," in Signal Processing and Communication Systems (ICSPCS), 2013 7th International Conference on, 2013, pp. 1 - 6.
  13. A. S. Rao, J. Gubbi, S. Marusic, and M. Palaniswami, "A robust algorithm for foreground extraction in crowded scenes," in 2012 International Symposium on Communications and Information Technologies, ISCIT 2012, 2012, pp. 604-609.
Blind-Aid Vision: In Australia, there are around 357,000 people who are blind or have low vision. Traditional white cane solution used by the elderly with vision impairment is inefficient as many are dependent on ambulatory aids, such as wheelchairs and walking frames. Uneven surfaces such as potholes and drop-offs cause non-protruding hazards. Current solutions based on proximity sensing is unsuitable for detecting non-protruding hazards. We have developed a proof-of-concept system to address this need. The system developed uses an optical laser system to identify the non-protruding hazards. It combines laser projection pattern, computer vision, pattern recognition, and machine learning. This was developed using a seed grant from the Ian Potter Foundation.

Details and Outcomes

Grants:

Collaborators: Tata Consultancy Services, India

Partners: Vision Australia; Guide Dogs, Australia

Outcomes: A portable prototype system for assisting disabled and elderly people that detects potholes and hazards (such as uneven surfaces). It uses optics, computer vision and machine learning techniques to deliver the results

Publications:

  1. A. S. Rao, J. Gubbi, M. Palaniswami, and E. Wong, "Non-Protruding Hazard Detection for the Aged Vision-Impaired," in 2016 IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS), 2016.
  2. A. S. Rao, J. Gubbi, M. Palaniswami, and E. Wong, "A vision-based system to detect potholes and uneven surfaces for assisting blind people," in Communications (ICC), 2016 IEEE International Conference on, 2016, pp. 1-6.

Media:

Biomedical Image Processing: Parkinson’s disease (PD) is a progressive neuro-degenerative disorder increasingly prevalent over the age of 65. In Australia, about 80,000 Australians suffer from PD. Changes in voice are amongst the earliest symptoms of the disease, preceding the motor manifestations by up to 9 years. Computed tomography (CT) images can be used to detect vocal fold motion changes related to these changes in voice. The anatomical position of the vocal folds is affected by the patient position, motion, phonation and breathing during scanning. Currently, radiologists manually adjust the vocal fold plane to analyse vocal cord changes manually. The large amount of modern CT imaging data capturing vocal fold plane makes it laborious and time-consuming. Localizing the vocal fold plane is a complex task and no known research automates the vocal fold plane extraction. Previously, we also worked to track tumour using PET images.

Details and Outcomes

Collaborators: Monash Medical Centre, Australia; Monash University, Australia; Peter MacCallum Cancer Institute, Australia; Royal Brisbane Hospital, Australia;

Outcomes: Automated image segmentation and tracking algorithms from biomedical images.

Publications:

  1. S. Hewavitharanage, J. Gubbi, D. Thyagarajan, K. Lau, and M. Palaniswami, "Automatic segmentation of the rima glottidis in 4D laryngeal CT scans in Parkinson's disease," in 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2015, pp. 739-742.
  2. S. Hewavitharanage, J. Gubbi, D. Thyagarajan, K. Lau, and M. Palaniswami, "Estimation of vocal fold plane in 3D CT images for diagnosis of vocal fold abnormalities," in 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2015, pp. 3105-3108.
  3. J. Gubbi, M. Palaniswami, K. Tomas, D. Binns, and M. Griffiths, "Delineation of region of interest volume in cardiac gated PET images," in Computers in Cardiology, 2009, pp. 765-768.
  4. A. Kanakatte, J. Gubbi, B. Srinivasan, N. Mani, T. Kron, D. Binns, et al., "Pulmonary tumor volume delineation in PET images using deformable models," in Proceedings of the 30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS'08 - "Personalized Healthcare through Technology", 2008, pp. 3118-3121.

Healthcare Research

Stroke Detection: Is one of leading cause of disability and mortality. An estimated 15 million people suffer from stroke incidence and around 6 million deaths every year. In 2015, an estimated 440,000 Australians and 1.6 million Indians suffered from stroke. Early detection of stroke is important to save people’s lives as late treatment will permanently damage brain causing death. Around 80% of cases are because of blood clots in brain. Administration of a thrombolytic medication to unblock the arteries after immediate onset (within 24 hours) of stroke is critical. However, administering thrombolytic agents require continuous monitoring. The first 24 hours after the first onset stroke has to be carefully monitored as stroke-like activities manifests and lack of continuous monitoring may lead to lost opportunities in timely treatments. In collaboration with Royal Melbourne Hospital, we have developed a proof-of-concept device based on accelerometer based on motor activity.

Details and Outcomes

Collaborators: Royal Melbourne Hospital, Australia

Outcomes: An algorithm based on accelerometry signals to detect affected arm and provide NIHSS score. A Stroke Watch has been designed to estimate the NIHSS score.

Publications:

  1. J. Gubbi, A. S. Rao, K. Fang, B. Yan, and M. Palaniswami, "Motor recovery monitoring using acceleration measurements in post acute stroke patients," BioMedical Engineering Online, vol. 12, 2013.
  2. J. Gubbi, D. Kumar, A. S. Rao, B. Yan, and M. Palaniswami, "A pilot study on the use of accelerometer sensors for monitoring post acute stroke patients," in Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS, 2013, pp. 957-960.
Epileptic Seizure Detection: Epilepsy is one of the most common neurological disorders. It is affecting about 50 million people worldwide. Patients with epilepsy are known to suffer from recurrent unprovoked seizures. One of the challenges with epilepsy is the unpredictability of the seizures. Seizures may cause significant risk on prolonged events and require emergency medical attention. We have developed a system that uses wrist-worn accelerometer. We have developed an algorithm to detect non-epileptic seizures using accelerometer data. The system uses machine learning to classify the seizures. This was developed in collaboration with Royal Melbourne Hospital.

Details and Outcomes

Collaborators: Royal Melbourne Hospital, Australia

Outcomes: Algorithms to detect convulsive and pseudo seizures based on unique accelerometry signatures.

Publications:

  1. J. Gubbi, S. Kusmakar, A. S. Rao, B. Yan, T. O'Brien, M. Palaniswami, et al., "Automatic Detection and Classification of Convulsive Psychogenic Nonepileptic Seizures Using a Wearable Device Classification of convulsive psychogenic non-epileptic seizures using muscle transforms obtained from accelerometry signal," in IEEE Journal of Biomedical and Health Informatics, 2016, pp. 1061-1072.
  2. S. Kusmakar, J. Gubbi, A. S. Rao, B. Yan, T. J. O'Brien, and M. Palaniswami, "Classification of convulsive psychogenic non-epileptic seizures using histogram of oriented motion of accelerometry signals," in Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS, 2015, pp. 586-589.
  3. J. Bayly, J. Carino, S. Petrovski, M. Smit, D. A. Fernando, A. Vinton, et al., "Time-frequency mapping of the rhythmic limb movements distinguishes convulsive epileptic from psychogenic nonepileptic seizures," Epilepsia, vol. 54, pp. 1402-1408, 2013.
Smartphone-based Low-cost Pulse Oximeter: Pulse oximetry is used in every developed-country clinic as a simple, painless way of measuring how much oxygen is in a patient's blood. Blood flowing into peripheral parts of the body (such as a fingertip or ear lobe) normally has almost 100% of the haemoglobin oxygenated, but in diseases such as pneumonia the "oxygen saturation" falls. An oximeter essentially consists of two LEDs, one emitting light in the red part of the spectrum and the other in the infrared, and a photo diode. However, despite their potential usefulness, oximeters are expensive (e.g. around $500) and rarely seen in developing countries. Cell phones are becoming increasingly common in developing countries such as Mozambique. The aim is to create a series of applications for cell phones as a platform for healthcare in remote and underserved communities in developing countries, using Mozambique as the site for development and testing.

Details and Outcomes

Collaborators and Partners: Microsoft; Nossal Institute for Global Health, Department of Information Systems, University of Melbourne

Outcomes: Designed prototype of a low-cost oximeter sensor and it is currently under clinical validation trials at western hospital, Victoria. It can connect to mobile phones to provide information about measured sensor values. We are examining applications for Australian remote communities.

Human Gait Pattern Analysis and Modelling: Understanding the underlying mechanisms and associated deficits in movement dynamics across the lifespan and the effects of pathological conditions, such as falls, will lead to many applications in the design and evaluation of diagnostic and assessment methods for human movement. Preliminary results obtained by our group have established the viability of the project for the "diagnosis" of gait instability and pathology. While in this application the experimental focus is both normal and ageing-influenced gaits, the outcomes are broad and generic having applications to other pathologies, and to rehabilitation, robotics and human performance. This work has been in collaboration with Victoria University Human Movement Lab. A few companies including SenSen Networks have expressed interest in such a technology apart from a few hospitals.

Details and Outcomes

Outcomes: Developed SVM-based machine learning models to detect and classify gait patterns.

Publications:

  1. D. T. H. Lai, P. T. Levinger, R. K. Begg, W. L. Gilleard, and M. Palaniswami, "Automatic recognition of gait patterns exhibiting patellofemoral pain syndrome using a support vector machine approach," IEEE Transactions on Information Technology in Biomedicine, vol. 13, pp. 810-817, 2009.
  2. D. T. H. Lai, R. K. Begg, and M. Palaniswami, "Computational intelligence in gait research: A perspective on current applications and future challenges," IEEE Transactions on Information Technology in Biomedicine, vol. 13, pp. 687-702, 2009.
  3. D. T. H. Lai, R. K. Begg, S. Taylor, and M. Palaniswami, "Detection of tripping gait patterns in the elderly using autoregressive features and support vector machines," Journal of Biomechanics, vol. 41, pp. 1762-1772, 2008.
  4. R. K. Begg, M. Palaniswami, and B. Owen, "Support vector machines for automated gait classification," IEEE Transactions on Biomedical Engineering, vol. 52, pp. 828-838, 2005.
Sleep Disordered Breathing: People with sleep disordered breathing (SDB) die prematurely with cardiovascular disease more often than someone without SDB. Early diagnosis and treatment of SDB could prevent adverse consequences. This investigation may provide essential information for introducing a novel screening device that can aid sleep specialist. This in turn may help prioritize patients, so that those in greatest need of treatment will undergo full PSG recordings in a timely manner, while those without apnoea will be able to avoid this tedious procedure. The work is supported by ARC Linkage grant scheme with Compumedics Ltd as the industry partner.

Details and Outcomes

Grants: ARC Linkage Projects (LP0454378) "New Techniques for Modelling, Diagnosis and Counter Measures for Cardiac Related Sleep Disordered Breathing"

Collaborator: Compumedics Ltd

Outcomes: We investigated into different machine learning techniques [support vector machines (SVM), Neural Network (NN), Quadratic discriminant (QD) model ] with an aim to find an appropriate model for the automatic detection of SAHS types from their respective overnight ECG recordings and estimation of relative degree of sleep disordered breathing.

Publications:

  1. C. Karmakar, A. Khandoker, T. Penzel, C. Schobel, and M. Palaniswami, "Detection of respiratory arousals using photoplethysmography (PPG) signal in sleep apnea patients," IEEE Journal of Biomedical and Health Informatics, vol. 18, pp. 1065-1073, 2014.
  2. A. H. Khandoker, C. K. Karmakar, T. Penzel, M. Glos, and M. Palaniswami, "Investigating relative respiratory effort signals during mixed sleep apnea using photoplethysmogram," Annals of Biomedical Engineering, vol. 41, pp. 2229-2236, 2013.
  3. J. Gubbi, A. Khandoker, and M. Palaniswami, "Classification of sleep apnea types using wavelet packet analysis of short-term ECG signals," Journal of Clinical Monitoring and Computing, vol. 26, pp. 1-11, 2012.
  4. A. H. Khandoker and M. Palaniswami, "Modeling respiratory movement signals during central and obstructive sleep apnea events using electrocardiogram," Annals of Biomedical Engineering, vol. 39, pp. 801-811, 2011.
  5. A. H. Khandoker, C. K. Karmakar, and M. Palaniswami, "Comparison of pulse rate variability with heart rate variability during obstructive sleep apnea," Medical Engineering and Physics, vol. 33, pp. 204-209, 2011.
  6. A. H. Khandoker, M. Palaniswami, and C. K. Karmakar, "Support vector machines for automated recognition of obstructive sleep apnea syndrome from ECG recordings," IEEE Transactions on Information Technology in Biomedicine, vol. 13, pp. 37-48, 2009.
  7. A. H. Khandoker, C. K. Karmakar, and M. Palaniswami, "Automated recognition of patients with obstructive sleep apnoea using wavelet-based features of electrocardiogram recordings," Computers in Biology and Medicine, vol. 39, pp. 88-96, 2009.
  8. A. H. Khandoker, J. Gubbi, and M. Palaniswami, "Automated scoring of obstructive sleep apnea and hypopnea events using short-term electrocardiogram recordings," IEEE Transactions on Information Technology in Biomedicine, vol. 13, pp. 1057-1067, 2009.
  9. J. Gubbi, A. Khandoker, and M. Palaniswami, "Classification of obstructive and central sleep apnea using wavelet packet analysis of ECG signals," in Computers in Cardiology, 2009, pp. 733-736.
  10. M. A. Daulatzai, A. H. Khandoker, C. K. Karmakar, M. Palaniswami, and N. Khan, "Characterization of chimeric surface submentalis EMG activity during hypopneas in obstructive sleep apnea patients," in TIC-STH'09: 2009 IEEE Toronto International Conference - Science and Technology for Humanity, 2009, pp. 782-788.
  11. M. A. Daulatzai, N. Khan, C. Karmakar, A. Khandoker, and M. Palaniswami, "Lateral decubitus posture during sleep: Sub-groups of obstructive sleep apnea patients - Therapeutic value of vertical position in OSA," in ISSNIP 2009 - Proceedings of 2009 5th International Conference on Intelligent Sensors, Sensor Networks and Information Processing, 2009, pp. 181-184.
  12. K. Nilsen, E. Zilberg, D. Burton, A. H. Khandoker, and M. Palaniswami, "Variations in the accuracy of the ECG based detection of obstructive sleep apnoea (OSA) for different numbers of ECG leads and categories of OSA events," in Proceedings of the 30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS'08 - "Personalized Healthcare through Technology", 2008, pp. 3492-3495.
  13. A. H. Khandoker, C. K. Karmakar, and M. Palaniswami, "Analysis of coherence between sleep EEG and ECG signals during and after obstructive sleep apnea events," in Proceedings of the 30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS'08 - "Personalized Healthcare through Technology", 2008, pp. 3876-3879.
  14. A. H. Khandoker, C. K. Karmakar, and M. Palaniswami, "Cross power spectral density between two-lead ECG signals at the termination of obstructive sleep apnea with or without arousals," in Computers in Cardiology, 2008, pp. 689-691.
  15. A. H. Khandoker, C. K. Karmakar, and M. Palaniswami, "Interaction between sleep EEG and ECG signals during and after obstructive sleep apnea events with or without arousals," in Computers in Cardiology, 2008, pp. 685-688.
  16. A. H. Khandoker, C. K. Karmakar, and M. Palaniswami, "Power spectral analysis for identifying the onset and termination of obstructive sleep apnoea events in ECG recordings," in Proceedings of ICECE 2008 - 5th International Conference on Electrical and Computer Engineering, 2008, pp. 96-100.
  17. A. H. Khandoker, C. K. Karmakar, and M. Palaniswami, "Analysis of coherence between sleep EEG and ECG signals during and after obstructive sleep apnea events," Conference proceedings : Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference, pp. 3876-3879, 2008.
  18. A. H. Khandoker, J. Gubbi, and M. Palaniswami, "Recognizing central and obstructive sleep apnea events from normal breathing events in ECG recordings," in Computers in Cardiology, 2008, pp. 681-684.
  19. C. K. Karmakar, A. H. Khandoker, and M. Palaniswami, "Identification of onset, maximum and termination of obstructive sleep apnoea events in single lead ECG recordings," in Proceedings of the 30th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS'08 - "Personalized Healthcare through Technology", 2008, pp. 1072-1075.
  20. T. Al-Ani, C. K. Karmakar, A. H. Khandoker, and M. Palaniswami, "Automatic recognition of obstructive sleep apnoea syndrome using power spectral analysis of electrocardiogram and hidden markov models," in ISSNIP 2008 - Proceedings of the 2008 International Conference on Intelligent Sensors, Sensor Networks and Information Processing, 2008, pp. 285-290.
  21. A. H. Khandoker, C. K. Karmakar, and M. Palaniswami, "Screening obstructive sleep apnoea syndrome from electrocardiogram recordings using support vector machines," in Computers in Cardiology, 2007, pp. 485-488.
  22. C. K. Karmakar, A. H. Khandoker, and M. Palaniswami, "Power spectral analysis of ECG signals during obstructive sleep apnoea hypopnoea epochs," in Proceedings of the 2007 International Conference on Intelligent Sensors, Sensor Networks and Information Processing, ISSNIP, 2007, pp. 573-576.

Wireless Sensor Networks Research

Urban Noise Monitoring: Excessive noise could potentially be harmful. Some of the frequencies are known to cause increased disturbances among people with autism. We investigated the problem to provide a solution based on wireless networks. The solution offers city councils to understand dynamic noise pollution profile, assess its impact on health and wellbeing, and better plan for noise reduction and desirable urban sound-scape. This solution was developed in collaboration with the City of Melbourne.

Details and Outcomes

Contract: Noise Mapping: Designing an Urban Information Architecture to Record and Map Noise Pollution.

Partner: City of Melbourne

Outcomes: We proposed a frameowrk to monitor noise in urban environments. We also eveloped low-cost noise monitoring hardware considering energy-efficiency of noise monitoring sensor network, and development of protocols for efficient collection of high-sample rate noise data from sensors deployed over large geographical regions. This also included developing methodologies for analysis and visualisation of the noise pollution profile.

Publication:

  1. J. Gubbi, S. Marusic, A. S. Rao, Y. W. Law, and M. Palaniswami, "A pilot study of urban noise monitoring architecture using wireless sensor networks," in Proceedings of the 2013 International Conference on Advances in Computing, Communications and Informatics, ICACCI 2013, 2013, pp. 1047-1052.

Media:

Air Quality Monitoring - Particulate Matter (PM): Increased level of pollution in the atmosphere contributes to respiratory health problems in the population. Particulate matter (PM), which is a fluid mixture of solids and liquids, is arguably one of the most dangerous of all pollutants to cause a population health hazard. Atmospheric pollutants, such as gases and particulate matters (PM) pose a threat to human health. In particular, there has been a strong focus on particulate matter as it is a common pollutant to cause population health hazards, especially respiratory illness. Monitoring of this pollutant is currently attained at low spatial resolutions due to the cost of accurate sensing devices. Even though these devices are highly accurate, given the distance they are placed apart from each other, the relevance of their measurements to an unmeasured spatial location in between sensors will be very low, which causes large estimation errors.

Details and Outcomes

Grants: ARC Linkage Projects (LP120100529); ARC LIEF (LE120100129); EU FP7 SocioTal and H2020 Organicity

Outcomes: We developed wireless sensor nodes equipped with low-cost PM sensors to supplement the existing high-accuracy PM devices to improve the estimation at higher spatial and temporal resolutions. Furthermore, we designed spatiotemporal estimation algorithms and tools to efficiently estimate the data at unobserved locations using the combination of high-cost and low-cost sensors.

Publications:

  1. S. Rajasegarar, T. C. Havens, S. Karunasekera, C. Leckie, J. C. Bezdek, M. Jamriska, et al., "High-resolution monitoring of atmospheric pollutants using a system of low-cost sensors," IEEE Transactions on Geoscience and Remote Sensing, vol. 52, pp. 3823-3832, 2014.
  2. S. Rajasegarar, P. Zhang, Y. Zhou, S. Karunasekera, C. Leckie, and M. Palaniswami, "High resolution spatio-temporal monitoring of air pollutants using wireless sensor networks," 2014 IEEE Ninth International Conference on Intelligent Sensors, Sensor Networks and Information Processing (ISSNIP),2014, pp. 1-6.
Environmental Monitoring:

Great Barrier Reef Ocean Observing System (GBROOS): The Great Barrier Reef (GBR) consists of 3200 coral reefs extended over 280,000km2. Understanding the patterns of thermal stress and other environmental parameters is essential for monitoring the health of the coral reefs (eg., refer [10] for environmental change and impacts on GBR). The health of the coral reefs can be affected by cold water intrusions, hot water intrusions, coral calcification, ocean acidification, coral-algae phase shifts and the spread of coral diseases due to temperature increases. One of the GBROOS aim is to install sensor networks at seven reef sites along the GBR and monitor the reefs.

Smart Environmental Monitoring and Analysis Technologies (SEMAT): The project aims to progress a working proof-of-concept next generation wireless marine sensor network to a commercial prototype to enable real time monitoring of the marine environment. The system will utilise an integrated, sensor network solution using underwater radio-frequency communication.

Sensor Map: Recent developments in technology together with widely observed climate change phenomena have revealed coral reef ecosystems as critical areas greatly susceptible to impact of global climate variations as well as other man-made influences, but also as early indicators of such events. The need to understand and protect such delicate ecosystems has created an urgent demand for the sensor networks technologies to be deployed in order to perform essential environmental monitoring and information collection. The project has created a facility to publish the sensed data (visual, temperature, light, etc.) directly on the World Wide Web. The online data published periodically can be accessed on MS SensorMap webpage from anywhere.

Details and Outcomes

GBROOS - Partners and Collaborators: Australian Institute of Marine Science (AIMS); James Cook University; University of Queensland;

SEMAT - Partners and Collaborators: University of Queensland, James Cook University, Queensland Cyber Infrastructure Foundation, Torino Foundation, Politechnic di Milan and the Danish Hydraulics Group Australia

SensorMap -Partners and Collaborators: Microsoft funded project (SensorMap for Great Barrier Reef); Australian Institute of Marine Science (AIMS); James Cook University

Outcomes: Developed sophisticated algorithms for modelling the effects of pollution in the Great Barrier Reef (GBR) from the deployed sensor networks. Our algorithms were able to detect the Cyclone Hamish (formed in the northern part of Australia and tracked through the main part of the reef) on the 9th of March 2009 using the sensor network stations at Heron and One Tree Islands.

Publications:

  1. A. Pirisi, F. Grimaccia, M. Mussetta, R. E. Zich, R. Johnstone, M. Palaniswami, et al., "Optimization of an energy harvesting buoy for coral reef monitoring," in 2013 IEEE Congress on Evolutionary Computation, 2013, pp. 629-634.
  2. S. Rajasegarar, J. Gubbi, O. Bondarenko, S. Kininmonth, S. Marusic, S. Bainbridge, et al., "Sensor network implementation challenges in the great barrier reef marine environment," in Proceedings of the ICT-MobileSummit 2008 Conference, 2008.

Additional Links:

  1. GBROOS
  2. SEMAT
  3. SensorMap
BigNet Testbed: The development of a world class national research and industry capability in sensor network technology is of great national importance. Although Wireless Sensor Networks (WSNs) are being studied on a global scale, current research is still focusing on simulations and small-scale experiments. In particular for sensor networks, which have to deal with very stringent resource limitations and that are exposed to severe physical conditions, real experiments with real applications are essential. The effectiveness of simulation studies is severely limited in terms of the lack of scalability of network simulators for large networks, as well as the difficulty in modelling the complexities of the radio environment, power consumption on sensors, and the interactions between the physical, network and application layers. Therefore, we propose to face the real world, extending the simulations and laboratory experiments to large scale, real applications. Only by doing so will WSN research: (1) get an insight into the real needs of the users of sensor network; (2) face the world of errors, incomplete information, dynamics, etc. that is very hard to model; (3) perform large-scale experiments of several hundreds of sensor nodes. In the BigNet project, we intend to use this experimental facility as a testbed to demonstrate the algorithms developed on a large scale, facing real world problems, and developing solutions for a variety of applications

Details and Outcomes

Grant:ARC LIEF Grant (LE0883073) - "Bignet - A Distributed Wireless Sensor Network Testbed"

Collaborators: The University of Melbourne; James Cook University; Deakin University

Outcomes - deployment projects:

  • Generic Sensor Network Testbed for Anomaly Detection
  • Wireless marine sensor networks
  • WSN Security
  • Wireless Multimedia Sensor Network – A case study
  • Wireless Body Area Networks
  • Wireless Sensing and Monitoring for Aged Healthcare
  • Spatial Computation in WSNs
  • Data-centric routing and collection
  • Web and grid enablement
  • Laboratory for Integrated sensing and networking for real time data acquisition and visualization (Deakin)
  • High Data Rate Marine Information Stream

Additional Link: