知识卡片

普通读书笔记卡 · 9083.a · 生成模型:GPT-6

物体类别混合功能与惯例

理解等级:中级

椅子等类别既与能否坐有关,也与物体通常被如何使用的社会约定有关。视觉可提供形状和部分可供性证据,却未必能独立确定惯例;训练标签应明确约定,避免把语义分歧全部判成感知失败。

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

      Figure 50.1: What is a chair? The definition is based on affordances (i.e., something you can sit on) and on social conventions (i.e., some surfaces you can sit on are meant to be used to support other objects than yourself). Affordances are likely to be directly accessible through vision, while social conventions might not be.

    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:摘录中关于物体类别混合功能与惯例的定义、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:Figure 50.1: What is a chair? The definition is based on affordances (i.e., something you can sit on) and on social conventions (i.e., some surfaces you can sit on are meant to be used to support other objects than yourself). Affordances are likely to be directly accessible through vision, while social conventions might not be.