Microsoft Research

Our Microsoft funded project investigates novel approaches in addressing this challenge. We were provided 7 SenseCams (passively capturing wearable cameras that take approximately 2,000 images on average per day). A large image collection will quickly build up, with an average of 650,000 images captured each year. This presents a considerable challenge in terms of managing such a large collection and to make it accessible for users.

Columbia University

We investigated the optimal combination of data sources towards activity detection, i.e. breaking up all the images into distinct events/activities. To realise this challenge we recorded 10 days of SenseCam data, as well as personal audio recorded from an MP3 player. Aiden Doherty spent a 1 week research visit with the LabROSA audio processing group of Prof. Dan Ellis, where our image and sensor analysis techniques were combined with the audio processing techniques of Columbia University. The collaboration resulted in a publication at the RIAO conference.

University of Amsterdam

Working with Cees Snoek of the Intelligent Systems Lab, Informatics Institute, we jointly developed a set of semantic concepts suitable for SenseCam images and applied these to a large dataset gathered from 5 different users. The SMAT 2008 paper (pdf) reports our progress on this and shows how semantic concepts can be used to filter SenseCam images in user browsing, and in other SenseCam applications.

CWI

Alia Amin, a PhD student at the Semantic Media Interfaces group, CWI, Amsterdam, is interested in using SenseCam as part of obtaining user behaviour data in establishing the requirements for designing new information systems, in order to complement the conventional ethnographic methods conducted at the beginning of a project, such as observation and interview.

Lulea University of Technology

Basel Kikhia, PhD student in Computer Science (Media Technology), Lulea University of Technology, Sweden, will visit DCU during April and May 2009 as part of the MemoryLane project.

Oliver Zanwill Centre, Cambridge

The Oliver Zangwill Centre for Neuropsychological Rehabilitation is part of Cambridgeshire Community Services which provides high quality rehabilitation for the individual cognitive, social, emotional and physical needs of people with acquired brain injury. Our work with the OZC has been on combining SenseCam and Actiheart, both of which act as lifelogging devices, but both of which work independently. The Actiheart, a combined heart rate and movement sensor, is designed to assess physical activity and calculate energy expenditure of it's wearers. The device is worn on the chest and is ideal for ambulatory monitoring. By integrating the outputs of the two devices we can get a much more complete picture of the wearer's activities during the day.

University of Leeds

Working with Dr. Chris Moulin and Prof. Martin A. Conway at the Institute of Psychological Sciences we are jointly conducting research on how the SenseCam aids recall of autobiographical memory. Early indications are that there may be some very exciting outcomes from this research.

University of Toronto

Prof. Ron Baecker of the Knowledge Media Design Institute is conducting research into investigating whether the SenseCam can help enhance the quality of life of people who have Alzheimer's Disease. Prof. Baecker's group are investigating connections between lifelogging and digital storytelling and have been making use of our event segmentation software to assist their analysis. Knowledge transfer is ongoing in this open collaboration.

University of Utrecht

During 2008 we worked with Wolfgang Huerst at the University of Utrecht, the Netherlands, where Wolfgang gathered his own SenseCam image collection for the purpose of using this as a testbed for his fast, mobile image viewing software. We are still in contact with Dr. Heurst and investigating future collaboration opportunities.

University of Illinois at Urbana Champaign

Prof. William F. Brewer in the Department of Psychology is investigating using the SenseCam as a tool to study the memory processes in autobiographical memory. As part of this collaboration we have assisted Prof. Brewer's group in the management of SenseCam images by providing our SenseCam browsing system.

University of Tampere, Department of Information Studies

This work, carried out during 2008 with Prof. Kalervo Jarvelin of the University of Tampere, involved assessing how a SenseCam can be used in ethnographic studies to record a subject's activities and then assist with cued recall in post-activity de-briefing. Specifically, medication professionals in a health centre in Tampere, Finland, wore a SenseCam as part of their daily work and after work they were presented with their day's activities via an event-based SenseCam image browser, developed by us. The interviewer then stepped through the day, locating instances of where the wearer had information needs and how these were ultimately satisfied, and the SenseCam images acted to stimulate the wearer's recall of the day.

University of Ulster