Sensecam in Chicago

I return, jetlagged and sleep-deprived, from Chicago where I have been attending the SenseCam 2009 Symposium. For those not in the know, SenseCam is a little device that Microsoft has been prototyping for some years, a small camera worn around the neck like a very naff black medallion, which records the quotidian details of daily life by regularly and automatically taking photographs through its fish-eye lens, and associating them with metadata generated by sensors. It’s a simple yet fascinating idea.

The Symposium was co-located with a big neuroscience conference, so several brainy types were there to explain their use of SenseCams in scientific research. Emma Berry has pioneered the use of SenseCams with Alzheimer’s patients, and has shown that if they review the images of a particular event (the SenseCam takes two or three pictures per minute, I believe), this helps ‘fix’ the memory in their minds much better than other memory aids, such as photographs consciously posed and taken with a conventional digital camera, or even diaries of the event that they have written. This is an astonishing result, because the SenseCam images are unposed, often blurry and usually uninteresting. Several other talks showed that the SenseCam is an extremely useful tool for exploring memory. The creativity of the neuroscience community in using SenseCams to explore a wide variety of theories shows that there is a huge demand for such tools.

Techies were out in force as well, particularly from Alan Smeaton’s group at Dublin City University. They have focused less on the medical side of things, and looked instead at lifelogging, the practice of capturing and storing details of everyday life in an unprecedentedly complete record, an idea whose most famous practitioner is Gordon Bell of Microsoft, whose MyLifeBits project sponsored SenseCam’s initial development. DCU’s Cathal Gurrin has worn one of these things for years, and now has 5,000,000 images – a huge multimedia search task. But he has an incredible record of the last three or four years – for example, he has a photo of his first meeting with his girlfriend (they met in a lift – ahhh, sweet!).

The reactions to the SenseCam are mixed. In the abstract, people hate the idea as an egregious invasion of privacy – just check out the comments to a recent New Scientist article on the device (amazing how many comments breached their terms of use). On the other hand, Cathal maintains that only twice in all his career as a lifelogger has he been challenged. Most people, reassured that the SenseCam has no microphone, are OK with it. So it may be that faced with an actual social situation with a lifelogger, the privacy threats seem less intimidating than they do on paper. Certainly I had no great problem being spied on by three or four people simultaneously (though of course ends up self-censoring – a parallel threat to liberty?). Could this be the killer app for the sousveillance movement of Steve Mann and David Brin? A creep’s charter? Or a new total situation awareness tool for the authorities and the military? Or just a benign and permanent link with one’s past?

The Symposium was fascinating for the breadth of research being reported over two days – not only in disciplinary terms, but also the numbers of directions the research was going. Is the SenseCam there to provoke associations with thoughts and feelings not captured by the camera, or to decide matters of fact? Does it support subjective autobiographical memory, or is it a means for sharing information, supporting communities and being a resource for the carers of people with memory problems? How should they be integrated with other memory supports? And – the most vexed question – does its value arise from the presentation of all the unfiltered output (as Emma Berry argued strongly, in her paper on ‘The secret of SenseCam’, you can never know which image or sequence of images will provide the vital association)? Or, rather, should we expect the output to be filtered, only the ‘best’ pictures presented?

We simply do not know how such devices will affect our memory and our social bonds with the past. Memory and truth have only recently had a strong connection (made strong by writing, printing, photography and film), and memory has often been as much a tool for creating an identity as a store of information ready to hold you to account. Could such tools turn memory from a friend to a policeman? It’s good to hear the truth about yourself sometimes (especially if your memory is failing), but do we want it, or need it, all the time?

The SenseCam is soon to be commercialised by Vicon under the name ViconRevue, and will be available for either scientific research or lifelogging. The debate will continue.

About the author
Kieron O'Hara is senior research fellow in the School of Electronics and Computer Science, University of Southampton. His interests lie in the philosophy and politics of technology, specifically the world wide web and semantic web, looking how the web enables us to organise knowledge and how that affects society.

See all posts by Kieron O'Hara
July 2013

Search this blog