← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 8834.b · 生成模型:GPT-6

人脸检测与分析任务不同

理解等级:初级

人脸检测判断区域中是否有人脸;人脸分析进一步估计姿态、身份等属性。任务不同,错误和后果也不同,不能用检测效果替身份识别或属性分类担保。评价时先说明输出含义与用途,再决定采用什么指标与分组。

这张卡片看不懂?

已有反馈 · 0 条

    还没有人提交反馈。

    在整书笔记中的位置

    原书出处 · 1 处

    • · 查看原书依据

      Facial detection asks whether there is a face in some image region, while facial analysis refers to measuring additional attributes of the face, including pose and identity. It is important to characterize the performance of facial analysis systems across demographic groups, as certain studies have done [265]. A large Face Recognition Vendor Test by the National Institute of Standards and Technology (NIST) [176] examined the

    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:摘录中关于人脸检测与分析任务不同的定义、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:Facial detection asks whether there is a face in some image region, while facial analysis refers to measuring additional attributes of the face, including pose and identity. It is important to characterize the performance of facial analysis systems across demographic groups, as certain studies have done [265]. A large Face Recognition Vendor Test by the National Institute of Standards and Technology (NIST) [176] examined the