知识卡片
视觉系统错误可转化为现实伤害
视觉系统进入身份、驾驶和公共空间后,预测错误的后果可能远超一个基准分数。需要把受影响对象、错误类型与使用方式一起评估。潜在社会收益并不自动抵消具体伤害,也不能因某任务可自动化就认为应该部署。
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
computer vision mistakes can have a large impact and have the potential to cause harm, and unfortunately this is true. While self-driving cars may save tens of thousands of lives each year, mistakes in those algorithms can also lead to human deaths. A mistaken computer match with a surveillance photograph may implicate the wrong person in a crime. Furthermore, the performance of computer vision algorithms
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:摘录中关于视觉系统错误可转化为现实伤害的定义、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:computer vision mistakes can have a large impact and have the potential to cause harm, and unfortunately this is true. While self-driving cars may save tens of thousands of lives each year, mistakes in those algorithms can also lead to human deaths. A mistaken computer match with a surveillance photograph may implicate the wrong person in a crime. Furthermore, the performance of computer vision algorithms