2018

13 June

Invited keynote on visual monitoring of people in AAL applications at BMVA Symposium: Computer Vision for smart environments and assisted living in London.

13 February

Presentation of our new EU project PAAL - Privacy-Aware and Acceptable Lifelogging services for older and frail people, funded under the framework of the Joint Programming Initiative "More Years, Better Lives," at the JPI MYBL Conference.

1 February

The Special Issue on Computer Vision in Healthcare and Assisted Living in the IET Computer Vision journal is published.

9 January

Paper Recognition of Activities of Daily Living Based on Environmental Analyses Using Audio Fingerprinting Techniques: A Systematic Review publised in the Sensors journal.

2017

11 October

Paper "Vide-omics: A Genomics-inspired Paradigm for Video Analysis" published in Computer Vision and Image Understanding journal.

30 June

Viva of PhD Thesis "Development of a scalable, reproducible and semantically interoperable emergency triage system" by my student José Miguel Bellod Guillén.

3 March

Looking for experts and companies interested/working in image-based trade mark retrieval. Contact me.

2016

15 December

Invited talk at the International Workshop on Assistive & Rehabilitation Technology (IWART 2016).

24 October

2nd IET International Conference on Technologies for Active and Assisted Living (TechAAL 2016) chaired in London.

24 October

Paper "Human Action Recognition Based on Temporal Pyramid of Key Poses Using RGB-D Sensors" presented by Enea Cippitelli at ACIVS 2016 is awarded the Best Student Paper prize.

15 October

Published call for papers on Computer Vision in Healthcare and Assisted Living for the IET Computer Vision journal. Deadline on 23 January 2017.

28 September

Workshop on "Expectations for AAL and Enhanced Living Environments" chaired at the AAL Forum 2016.

4 September

Paper "Validation Techniques for Sensor Data in Mobile Health Applications" has been published in the Journal of Sensors.

18 August

Book "Active and Assisted Living: Technologies and Applications" published by The Institution of Engineering and Technology.

13 August

Paper "Weakly supervised activity analysis with spatio-temporal localisation" published in Neurocomputing journal.

19 July

Paper "Action identification using a descriptor with autonomous fragments in a multilevel prediction scheme" published in Signal, Image and Video Processing journal.

14 June

Published call for papers for the 2nd IET International Conference on Technologies for Active and Assisted Living (TechAAL 2016).

3 June

Paper Identification of Activities of Daily Living Using Sensors Available in off-the-shelf Mobile Devices: Research and Hypothesis presented at the 7th International Symposium on Ambient Intelligence (ISAmI 2016).

1 June

Paper Kinect and Episodic Reasoning for Human Action Recognition presented at the 13th International Conference on Distributed Computing and Artificial Intelligence (DCAI 2016).

9 March

Discussion paper The ethics of using cameras in care homes published in Nursing Times.

12 February

I offer 3 funded PhD opportunities: "Towards a robust method for person segmentation in videos. Application to ensure privacy in active and assisted living services", "Parallel evolutionary algorithm for dense pixel matching in presence of distortions" and "Evolutionary ordering of multidimensional spaces. Application to image processing and computer vision". Deadline: 18/03/16. How to apply and details. Look for them in the list. Please, you have to state that you apply to the project you are interested in.

If you intend to apply for a Studentship and are interested in one of the projects below, the outline proposal listed for each project will satisfy the requirement to submit a project proposal as part of your application. So, there is no need to submit now a complete project proposal.

Towards a robust method for person segmentation in videos. Application to ensure privacy in active and assisted living services

In the last years, Kingston University has participated in different projects (caring4U and BREATHE) to design and develop video-based systems to monitor people in private spaces. The main application of these systems have focused on active and assisted living (AAL). AAL refers systems aimed at improving the quality of life and supporting independent and healthy living of older or impaired people by using information and communication technologies (ICT) at home. However, this type of monitoring using cameras can be seen as intrusive and as violating rights to privacy, because of the concern that raw video images could be viewed by unauthorised viewers or stored for inappropriate use.

Therefore, we have been working in different mechanisms to ensure privacy in videos by detecting the person/s in the scene and modifying the image so as to increase privacy while maintaining the necessary intelligibility required for the applications. This methods have been employed with both RGB cameras and RGB-D devices (Kinect or similar).

Different filters that can be applied to protect privacy of a person

The first stage in these methods is to locate the person and segmented from the background. While this is easier with RGB-D devices, it is still very complicated to achieve good results with RGB cameras under real conditions, where shadows, changes in lighting, occlusions due to furniture, may affect the segmentation.

The aim of this project is to design and develop a robust method for the segmentation of people in indoor spaces. There are many algorithms for background extraction/foreground detection. However, these algorithms work differently according to the conditions of the environment. Therefore, the idea is to develop a method that decides at every moment which is the appropriate algorithm or combinations of them in order to improve the segmentation. The method will also autonomously decide when the algorithm needs to be changed or re-started. It will also considered machine learning methods able to recognise if a person is in the scene.

Evolutionary ordering of multidimensional spaces. Application to image processing and computer vision

Mathematical morphology (MM) is used for image processing in different applications: noise filtering, shape simplification, edge detection, skeletonisation, shape analysis, segmentation... MM has two basic operators, erosion and dilation, from which many other morphological operations are derived. These operators rely on an ordering of the pixel values. MM is applied to a set provided with an order and with a supremum and an infimum pertaining to that order. This is the reason why MM has been mainly applied to binary or grayscale images. Extension to multivariate data, as colour images, is not straightforward, because there is not an ordered relationship among vectors.

This project will investigate different alternatives to obtain orderings that preserve the topology of the input space. Besides, as calculating those orderings are computationally expensive, different proposals using evolutionary computation will be considered.

These evolutionary ordering methods will be applied to different of image processing and computer vision problems, considering a variety of multidimensional spaces: colours, histograms, bags of words...

Parallel evolutionary algorithm for dense pixel matching in presence of distortions

Dense pixel matching is an essential step required by many computer vision applications. While a large body of work has addressed quite successfully the rectified scenario, accurate pixel correspondence between an image and a distorted version remains very challenging.

This project will investigate the use of a new parallel evolutionary algorithm (EA) to find distortions between images. The structure of this EA replicates the topology of the problem so as good solutions for local distortions are propagated to neighbouring regions, obtaining the global distortion. A first approach of this idea was recently presented obtaining the best paper award. This project will improve the proposed idea validating it in different problems: stereo matching, tracking in medical images, support to visually­ impaired people...

2 February

Paper "From Data Acquisition to Data Fusion: A Comprehensive Review and a Roadmap for the Identification of Activities of Daily Living Using Mobile Devices" published in Sensors journal.

12 January

Paper "Automatic parameter tuning for functional regionalization methods" published in Papers in Regional Science.

7 January

Paper "Recognition of Activities of Daily Living with Egocentric Vision: A Review" published in Sensors journal.

2015

24 November

Invited talk "Visual monitoring of people in private spaces. From the "big brother" to the "good brother"" given at the Royal Society of Medicine.

5 November

Organisation of the IET International Conference on Technologies for Active and Assisted Living (TechAAL 2015).

16 October

My PhD student José Ramón Padilla López defends his doctoral thesis "Protección de la privacidad visual basada en el reconocimiento del contexto".

10 October

Lecture given in training school organised by COST Action IC1206 - De-identification for privacy protection in multimedia content.

24 September

Special session on "Monitoring people in private spaces: technological advances and societal issues" organised at the AAL Forum 2015.

19 August

Invited to give talk "Visual monitoring of people in private spaces. From the "big brother" to the "good brother"" at the Royal Society of Medicine.

16 July

Paper "Marginalised Stacked Denoising Autoencoders for Robust Representation of Real-Time Multi-View Action Recognition" published in Sensors journal.

12 June

Call for abstracts - Session on "Monitoring people in private spaces: technological advances and societal issues" at the AAL Forum 2015 - Deadline: 15th July.

4 June

Paper "Visual Privacy by Context: Proposal and Evaluation of a Level-Based Visualisation Scheme" published in Sensors journal.

26 May

Call for papers - IET International Conference on Technologies for Active and Assisted Living - Deadline: 4th September.

6 May

Call for abstracts - 3rd International Symposium SRUK/CERU Society of Spanish Researchers in the United Kingdom - Deadline: 30th May.

26 March

Call for papers - Computational Intelligence and Neuroscience journal, Special Issue on Computational Intelligence for Ambient Assisted Living.

20 February

I offer 2 funded PhD opportunities: "Parallel evolutionary algorithm for dense pixel matching in presence of distortions" and "Evolutionary ordering of multidimensional spaces. Application to image processing and computer vision". Deadline: 20/03/15. How to apply and details. Look for them in the list. Please, you have to state that you apply to the project you are interested in.

If you intend to apply for a Studentship and are interested in one of the projects below, the outline proposal listed for each project will satisfy the requirement to submit a project proposal as part of your application. So, there is no need to submit now a complete project proposal.

Parallel evolutionary algorithm for dense pixel matching in presence of distortions

Dense pixel matching is an essential step required by many computer vision applications. While a large body of work has addressed quite successfully the rectified scenario, accurate pixel correspondence between an image and a distorted version remains very challenging.

This project will investigate the use of a new parallel evolutionary algorithm (EA) to find distortions between images. The structure of this EA replicates the topology of the problem so as good solutions for local distortions are propagated to neighbouring regions, obtaining the global distortion. A first approach of this idea was recently presented obtaining the best paper award. This project will improve the proposed idea validating it in different problems: stereo matching, tracking in medical images, support to visually­ impaired people...

Evolutionary ordering of multidimensional spaces. Application to image processing and computer vision

Mathematical morphology (MM) is used for image processing in different applications: noise filtering, shape simplification, edge detection, skeletonisation, shape analysis, segmentation... MM has two basic operators, erosion and dilation, from which many other morphological operations are derived. These operators rely on an ordering of the pixel values. MM is applied to a set provided with an order and with a supremum and an infimum pertaining to that order. This is the reason why MM has been mainly applied to binary or grayscale images. Extension to multivariate data, as colour images, is not straightforward, because there is not an ordered relationship among vectors.

This project will investigate different alternatives to obtain orderings that preserve the topology of the input space. Besides, as calculating those orderings are computationally expensive, different proposals using evolutionary computation will be considered.

These evolutionary ordering methods will be applied to different of image processing and computer vision problems, considering a variety of multidimensional spaces: colours, histograms, bags of words...

11 February

Paper "Abnormal Gait Detection with RGB-D Devices using Joint Motion History Features" accepted at 1st International Workshop on Understanding Human Activities through 3D Sensors (UHA3DS'15)

2 February

Paper "Visual Privacy Protection Methods: A Survey" published in the Expert Systems with Applications journal

10 January

Call for chapters - Book on "Ambient Assisted Living: Technologies and Applications", edited by IET. Deadline for chapter proposals: 25/02/15

9 January

Paper "Topology-preserving ordering of the RGB space with an evolutionary algorithm" accepted at 18th European Conference on the Applications of Evolutionary Computation (EvoApps)

5 January

Paper "Measurement of Heel-Rise Test Results Using a Mobile Device" accepted at the Doctoral Consortium of the 2nd International Conference on Physiological Computing Systems (PhyCS)

2014

19 December

Talk given at Networking Workshop on "Salud y tecnologías aplicadas a la salud" organised by the Valencian Universities Network for the Promotion of Research, Development and Innovation (RUVID) and the Society of Spanish Researchers in the United Kingdom.

29 October

Paper "A Low-Dimensional Radial Silhouette-Based Feature for Fast Human Action Recognition Fusing Multiple Views" published in the International Scholarly Research Notices

30 September

Source code for the bag-of-key-poses machine learning method released

20 August

Papers "Visual Privacy by Context: A Level-Based Visualisation Scheme", "A Multiple Kernel Learning Based Fusion Framework for Real-Time Multi-View Action Recognition" and "Using wavelet transform to disaggregate electrical power consumption into the major end-uses" accepted at the 8th International Conference on Ubiquitous Computing & Ambient Intelligence (UCAmI) and 6th International Work-conference on Ambient Assisted Living (IWAAL 2014)

30 July

Paper "Continuous Human Action Recognition in Ambient Assisted Living Scenarios" accepted at the International Workshop on Enhanced Living Environments (ELEMENT-2014)

9 June

I offer 2 funded PhD opportunities: "Parallel evolutionary algorithm for dense pixel matching in presence of distortions" and "Recognition of activities of daily living with wearable vision systems". Deadline: 20/06/14. How to apply and details. Look for them in the list. Please, you have to state that you apply to the project you are interested in.

Parallel evolutionary algorithm for dense pixel matching in presence of distortions

Dense pixel matching is an essential step required by many computer vision applications. While a large body of work has addressed quite successfully the rectified scenario, accurate pixel correspondence between an image and a distorted version remains very challenging.

This project will investigate the use of a new parallel evolutionary algorithm (EA) to find distortions between images. The structure of this EA replicates the topology of the problem so as good solutions for local distortions are propagated to neighbouring regions, obtaining the global distortion. A first approach of this idea was recently presented obtaining the best paper award. This project will improve the proposed idea validating it in different problems: stereo matching, tracking in medical images, support to visually­ impaired people...

Recognition of activities of daily living with wearable vision systems

Ambient assisted living applications require real time monitoring of the state of a person. Cameras have only been recently considered in order to recognise activities of daily living as they provide rich information about what a person is doing. Recent project make use of different types of vision devices, e.g. RGB cameras, omnidirectional cameras, RGB­-D devices (e.g. Microsoft Kinect), located in the environment. Recognition of human activity at home is complex because of the possible cluttered environment and the occlusions created by the body, so the hands and interactions with objects are difficult to see.

Recent developments of wearable vision systems, e.g. Google Glass, may help to solve these problems. Therefore, this project will investigate the use of wearable vision systems to recognise human activity for ambient assisted living applications., e.g., mild cognitive impairment.

20 May

Invited speaker in seminar on vision-based monitoring of people in private spaces to the Sphere IRC at the University of Bristol

6 May

Paper "A Vision-based System for Intelligent Monitoring: Human Behaviour Analysis and Privacy by Context" accepted for publication at the Sensors journal

25 April

Our paper "Evolutionary algorithm for dense pixel matching in presence of distortions" gets the Best Paper Award of the Evolutionary computation in image analysis, signal processing and pattern recognition track at the 16th European Conference on the Applications of Evolutionary and bio-inspired Computation

15 April

Paper "Adaptive Human Action Recognition With an Evolving Bag of Key Poses" published at the IEEE Transactions on Autonomous Mental Development

16 February

Call for papers - European Conference on Artificial Intelligence - Workshop on "Artificial Intelligence for Ambient Assisted Living (AI4AAL)

14 February

Call for papers - IEEE Intelligent Systems, Special Issue on "Ambient Assisted Living and Robotics"

22 January

Paper "Evolutionary algorithm for dense pixel matching in presence of distortions" accepted in evostar 2014

2013

9 December

Encuesta sobre monitorización de personas y privacidad en el hogar

1 November

Paper "Optimizing Human Action Recognition based on a Cooperative Coevolutionary Algorithm" published at the Engineering Applications of Artificial Intelligence journal

8 October

Call for carers for BREATHE project

7 October

Paper "Fusion of Skeletal and Silhouette-based Features for Human Action Recognition with RGB-D Devices" accepted in 3rd IEEE Workshop on Consumer Depth Cameras for Computer Vision

5 October

Call for papers - IEEE Transactions on Human-Machine Systems, Special Issue on "Ambient Assisted Living - sensors, methods and applications"

22 August

Paper "Evolutionary joint selection to improve human action recognition with RGB-D devices" published in Expert System with Applications

16 August

Paper "Vision System for Monitoring the Activity of Daily Living at Home" accepted in IWAAL 2013

1 August

Announcement of side event AAL systems and associated privacy and ethical issues at the AAL Forum 2013

3 July

Paper "DAI Virtual Lab: a Virtual Laboratory for Testing Ambient Intelligence Digital Services" accepted in DRT4ALL 2013

1 May

BREATHE project starts

14 March

Paper "Human Action Recognition Optimization Based on Evolutionary Feature Subset Selection" accepted in GECCO 2013

28 February

Elevated to IEEE Senior Member

9 February

Paper "Silhouette-based Human Action Recognition using Sequences of Key Poses" accepted for publication in Pattern Recognition Letters journal

Frontpage Book on Technologies for Active and Assisted Living

Main projects

PAAL website BREATHE website caring4u website
metalTIC-Digital Home Webpage DAI Virtual Lab Webpage

Social networks

Profile in Facebook Profile in LinkedIn Profile in Twitter

Academic networks

Profile in Google Scholar Profile in DBLP Profile in Research Gate
Profile in Slideshare