Type

Conference Proceedings

Authors

Noel E. O'Connor
Cathal Gurrin
Alan F. Smeaton
Hyowon Lee
Aiden R. Doherty
Anil Kumar
Philip Kelly

Subjects

Computer Science

Topics
image retrieval visualization image classification lifelog interactive computer systems high resolution multimedia systems image processing

The colour of life: interacting with SenseCam images on large multi-touch display walls (2010)

Abstract A SenseCam can provide a detailed visual archive of a person’s life, activities and experiences. However, as the number of images captured per year can extend beyond one million, gaining an insight into an individual’s lifestyle in a fast, effective and intuitive manner is a challenging prospect. In this work, we develop an interactive image browsing tool, which incorporates visualisation techniques that can capture not only a snapshot of an individual’s lifestyle over long periods of time, but also how that lifestyle varies with changing days, weeks, or years. The image retrieval tool incorporates the Colour of Life algorithms [1], which can represent an overview of millions of images with a single visualisation. The Colour of Life algorithms focus on the relationship between lifestyle and colour, by capturing the colours to which we are exposed in our lives (and therefore captured by SenseCam images), collating similar colours for specific time periods and depicting how those colours change over time with a flowing time-line – see Figure 1 which depicts the life of a SenseCam user over the period of 8 days. In this figure, time is orientated along the horizontal axis and larger vertical peaks indicate higher user activity for a given period of time. In Figure 1, the normal working week consists of the rhythmical blue, pink (work) and yellow (home) peaks and troughs for each day (with less activity at the start and end of the days), whereas time outdoors increases at the weekend, especially during the night (and hence the darker colours on the left hand side of the figure). The Colour of Life visualisation, while providing information on changes in lifestyle, does not provide sufficient context to understand the exact activities of a user for a given time period. For example, on the left of Figure 1 there is a peak of purple, that does not occur anywhere else during the 8 days of activities images – where was the user at this point in time and what was he doing? In this work, we build an interactive image browsing tool based around the Colour of Life visualisation. We exploit the use of high resolution multi-touch display walls, where we extend the Colour of Life algorithms to produce an intuitive visualisation, which incorporates image mosaicing (see Figure 2). Through this we incorporate coarse lifestyle data with more fine detailed contextual information on human activities into one interactive visualisation tool. As an additional feature, we have investigated the use of image classification within the framework of the Colour of Life. One such example is the categorisation of images as being as social (i.e. interacting with other people) or non-social. Using such a classification, we can depict a person’s social lifestyle, and how that varies over time.
Collections Ireland -> Dublin City University -> DCU Faculties and Centres = Research Initiatives and Centres: Centre for Digital Video Processing (CDVP)
Ireland -> Dublin City University -> Status = Published
Ireland -> Dublin City University -> Subject = Computer Science: Multimedia systems
Ireland -> Dublin City University -> DCU Faculties and Centres = DCU Faculties and Schools: Faculty of Engineering and Computing: School of Electronic Engineering
Ireland -> Dublin City University -> DCU Faculties and Centres = Research Initiatives and Centres
Ireland -> Dublin City University -> Publication Type = Conference or Workshop Item
Ireland -> Dublin City University -> DCU Faculties and Centres = DCU Faculties and Schools: Faculty of Engineering and Computing: School of Computing
Ireland -> Dublin City University -> Subject = Computer Science: Interactive computer systems
Ireland -> Dublin City University -> Subject = Computer Science: Image processing
Ireland -> Dublin City University -> Subject = Computer Science
Ireland -> Dublin City University -> DCU Faculties and Centres = DCU Faculties and Schools
Ireland -> Dublin City University -> Subject = Computer Science: Lifelog
Ireland -> Dublin City University -> DCU Faculties and Centres = Research Initiatives and Centres: CLARITY: The Centre for Sensor Web Technologies
Ireland -> Dublin City University -> Subject = Computer Science: Visualization
Ireland -> Dublin City University -> DCU Faculties and Centres = DCU Faculties and Schools: Faculty of Engineering and Computing

Full list of authors on original publication

Noel E. O'Connor, Cathal Gurrin, Alan F. Smeaton, Hyowon Lee, Aiden R. Doherty, Anil Kumar, Philip Kelly

Experts in our system

1
Noel E. O'Connor
Dublin City University
Total Publications: 420
 
2
Cathal Gurrin
Dublin City University
Total Publications: 182
 
3
Alan F. Smeaton
Dublin City University
Total Publications: 450
 
4
Aiden R. Doherty
Dublin City University
Total Publications: 67