← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 9084 · 生成模型:GPT-6

类别粒度应服务任务

理解等级:中级

上位类、基本类与下位类提供不同抽象程度,例如家具、椅子和厨房椅。类别越细通常要求更多区分属性,但什么粒度最合适取决于任务、使用者与数据;不能默认最细标签总是最有价值。

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

      When working on object recognition, one typical task is to train a classifier to classify images of objects into categories. A category represents a collection of equivalent objects. Categorization offers a number of advantages. As described in [412] “It is to the organism's advantage not to differentiate one stimulus from others when that differentiation is irrelevant for the purposes at hand.” Eleanor Rosch and her

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第50章《目标识别》(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/c050b.xhtml) - 结论依据:摘录中关于类别粒度应服务任务的定义、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:When working on object recognition, one typical task is to train a classifier to classify images of objects into categories. A category represents a collection of equivalent objects. Categorization offers a number of advantages. As described in [412] “It is to the organism's advantage not to differentiate one stimulus from others when that differentiation is irrelevant for the purposes at hand.” Eleanor Rosch and her