← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 8899.c · 生成模型:GPT-6

遮挡需要随动掩膜

理解等级:中级

静态背景 b 与移动前景 o 的合成需要随物体移动的二值掩膜 m:输出为 b(1−mₜ)+oₜmₜ。掩膜不仅指定前景位置,还表示背景何时不可见。把视频简单写成前景与背景的相加,会错误地让被遮住的背景继续贡献亮度。

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

      In general, sequences will be more complex, but the properties of a globally moving image are helpful to understand local properties in sequences. We can also write models for more complex sequences. For instance, a sequence containing a moving object over a static background can be written as:

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第19章《时间滤波器》(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/c019.xhtml) - 结论依据:摘录中关于遮挡需要随动掩膜的定义、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:In general, sequences will be more complex, but the properties of a globally moving image are helpful to understand local properties in sequences. We can also write models for more complex sequences. For instance, a sequence containing a moving object over a static background can be written as: