当前位置: X-MOL 学术Sociology of Race and Ethnicity › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Interrogating Innovation
Sociology of Race and Ethnicity ( IF 1.8 ) Pub Date : 2020-08-06 , DOI: 10.1177/2332649220942519
Anne Pollock 1
Affiliation  

is filled predominantly with minority ethnic faces (likely to be the case given the U.S. criminal justice system’s longstanding predilection for ethnic minority incarceration), your face is probably less likely to be recognized if you are white. Additionally, if you are black, your face potentially stands a higher chance of being falsely recognized because of the limitations of facial recognition software that are understood to work more poorly with black faces. This is the likely result of a photographic industry so obsessed with representing whiteness that it neglected to develop adequate techniques for representing anything else. This last point is fantastically rendered in chapter three, “Coded Exposure,” where Benjamin situates this complexity within the broader history of visual technologies, including the Kodak color cards first used in the 1950s to send to photo labs to ensure the coloration of prints was correctly calibrated. The fact that the women featured on these cards were always white would set the standard for photography. Only later, when the profitability of overseas markets became apparent, would the color cards feature women of color. It is through this impressive tying together of history and the present that the potentially expansive scope of race critical code studies comes into view. Benjamin situates technologies like facial recognition AI within the broader sociomaterial structures that rendered it technically possible and that now sustain it. The software is fed with data that mirror the inequities of the social world, reproducing those very inequities. But while we could call a person a racist, we could never describe a computer as such—could we?

中文翻译:

质疑创新

由于主要是少数民族面孔(鉴于美国刑事司法系统长期以来对少数民族监禁的偏爱),如果您是白人,您的面孔可能不太可能被识别。此外,如果您是黑色的,则由于面部识别软件的局限性(被认为对黑脸的使用效果较差),您的脸部很可能会被错误识别。这可能是摄影行业如此痴迷于表现白色的结果,以至于它忽略了开发足够的技术来表现其他事物的可能性。最后一点在第三章“编码曝光”中得到了梦幻般的呈现,本杰明将这种复杂性置于更广阔的视觉技术历史中,包括1950年代首次用于发送给照相馆的柯达色卡,以确保正确校准打印件的色彩。这些卡片上的女性总是白人,这一事实将为摄影设定标准。直到后来,当海外市场的盈利能力变得明显时,色卡才会出现有色女性的形象。正是通过这种令人印象深刻的历史和现在的联系,种族关键代码研究的潜在范围才得以显现。本杰明将诸如面部识别AI之类的技术置于更广泛的社会材料结构中,这使其在技术上成为可能,并且现在得以维持。该软件提供了反映社会世界不平等现象的数据,从而再现了这些不平等现象。但是,尽管我们可以称呼某人为种族主义者,
更新日期:2020-08-06
down
wechat
bug