知识卡片
自动化偏见会加重纠错困难
人若倾向相信机器输出,被错误识别者可能更难纠正结论。部署流程需要考虑人怎样使用结果与处理异议,不能以有人审核就假设风险消失。模型评价和使用者行为共同影响实际后果。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
- · 计算机视觉与社会
原书出处 · 1 处
- · 查看原书依据
People can show a bias toward believing the output of a machine [87], making it more difficult for a person to correct an algorithm's mistake. There are concerns whether labeling the gender of a person from their appearance could cause distress or harm to some in the LGBTQ community [181, 45]. We provide a subsequent list of questions for thought and discussion. We encourage computer
参考来源
- 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第04章《计算机视觉与社会》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)/_epub-src/OPS/xhtml/c004.xhtml)
- 结论依据:摘录中关于自动化偏见会加重纠错困难的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。
- 原始内容:People can show a bias toward believing the output of a machine [87], making it more difficult for a person to correct an algorithm's mistake. There are concerns whether labeling the gender of a person from their appearance could cause distress or harm to some in the LGBTQ community [181, 45]. We provide a subsequent list of questions for thought and discussion. We encourage computer