知识卡片

普通读书笔记卡 · 8978.f · 生成模型:GPT-6

顺序深度不同于度量深度

理解等级:中级

纸堆各页的间距可能很小,仍可根据遮挡判断谁在上面。层表示记录这种前后次序,即使不知道具体距离也能支持取物等操作。度量深度回答相距多少,顺序结构回答谁遮挡谁;二者信息不同,也不保证在复杂交叉场景中存在单一全局层序。

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

      Layers are not just an approximation to true depth, but also capture geometric structure that is not well-represented by depth. Consider a pile of papers on your desk. The depth difference between one page and the next might only be a millimeter—too small to show up on standard depth sensing cameras. But the layer structure is obvious. Look back at figure 31.2: the depth order

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第31章《知觉分组》(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/c031.xhtml) - 结论依据:摘录中关于顺序深度不同于度量深度的定义、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:Layers are not just an approximation to true depth, but also capture geometric structure that is not well-represented by depth. Consider a pile of papers on your desk. The depth difference between one page and the next might only be a millimeter—too small to show up on standard depth sensing cameras. But the layer structure is obvious. Look back at figure 31.2: the depth order