The ‘racist’ webcam: Anthropometry, design and the myth of objective science

Posted by Ian on Dec 22, 2009 in blog | No Comments
A You Tube video posted by wzamen01 on 10 December 2009 demonstrates how the new face tracking software of HP’s new webcams is unable to track a black man’s face. Admirably, the video post is made in good humour and is a positive example of effective consumer criticism, but what I am most concerned with are the comments and responses to the videos that have been made.

In response to the posting, HP released this statement:
‘We are working with our partners to learn more. The technology we use is built on standard algorithms that measure the difference in intensity of contrast between the eyes and the upper cheek and nose. We believe that the camera might have difficulty “seeing” contrast in conditions where there is insufficient foreground lighting.’
Although this might be a technically appropriate explanation of the issue being experienced, it is revealing that the product was released by HP without this having been considered or at least brought to user/public attention.
Just this morning on the You Tube, user BaldEspresso posted a comment to the video stating that:
‘Physics cannot see race. Only colour. I’ll assume this video was made as a joke.’
This comment is technically correct in that physics (or other sciences) works with objective/quantifiable concepts such as colour, weight, etc. and that subjective concepts such as race can not be directly measured. However, the personification of physics as ‘seeing’ anything in the sentence is ironic. If we’re being technical, physics does not see anything. It’s not alive. Or is it…
Science is more than numbers and data. It is a set of relations, observations and agreements between people in the world. Using a fundamentalist definition of science that is above human agency ignores not only the critical issues of design and application of systems (which is as much a part of science as the numbers) as well as the long history of scientific racism of the last two centuries.
Head-measurer tools of the late 19th and early 20th centuries were only intended to measure heads. Humans made the intuitive leaps to develop thought-systems and hierarchies of anthropometrics (human measurement) to justify racism and genocide. So it’s not at issue whether HP’s technology is itself racist. Rather, it is whether the applications of the technology and the assumptions about users made at the beginning of the design process that may reveal oversights or ignorance.
Many companies use anthropometric data in designing their products today (such as clothes manufactures, car companies and NASA). This is to optimise the experience or use of a product for the largest number of users. Every country in the world tends to develop their own sets of anthropometric data that is revised regularly to account for fluctuations in height, weight, size, etc. due to changes in lifestyles, diets, etc.
The question to be asked today of HP is what sort of anthropometric data did they utilise to account for skin tone colour variation amongst the world’s population and who compiled this data? Or whether they considered any such data set and merely made assumptions about user face colour tone and contrast which would suggest possible indirect racism on behalf of the designers and programmers.
The lesson from this situation are that science is not above human bias, ignorance or flaw. Science is not an omnipotent being distinct from humans above arguments over subjective values. Science is deeply embedded within humanity, and our application of scientific thought and data in developing technologies must always be reflexively aware of this. Furthermore, the process of design and the development of technology and the programming of software are also not simply separate, functional, objective pursuits, and perhaps this failure in HP’s software and technology today reveals an unhealthy distance that has been growing between software designers and programmers and humanity – a same distance which unfortunately grows between the sciences and the arts and humanities.

EmailBufferFacebookTwitterTumblrLinkedInGoogle+PinterestSina Weibo

Leave a Reply