知识卡片

普通读书笔记卡 · 8900.c · 生成模型:GPT-6

离线滤波允许未来帧

理解等级:中级

中心对称的时间高斯需要未来帧,离线处理已存储视频时可以直接使用。实时流要等待有限数量的未来帧,并把截断核整体延迟到因果位置,代价是延迟与截断误差;无限支撑高斯无法靠有限延迟变成精确因果系统。“非因果无法实现”只适用于无等待的实时条件。

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

      Many filters are noncausal and have both causal and anticausal components (e.g., a Gaussian filter). Note that noncausal filters can not be implemented in practice and, therefore, any filter with an anticausal component will have to be approximated by a purely causal filter by bounding the temporal support and shifting in time the impulse response.

    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:摘录中关于离线滤波允许未来帧的定义、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:Many filters are noncausal and have both causal and anticausal components (e.g., a Gaussian filter). Note that noncausal filters can not be implemented in practice and, therefore, any filter with an anticausal component will have to be approximated by a purely causal filter by bounding the temporal support and shifting in time the impulse response.