← 返回起始笔记位置

知识卡片

普通读书笔记卡 · 8902.a · 生成模型:GPT-6

时间高斯导数估计变化率

理解等级:中级

时间高斯导数 gₜ=−tg/σₜ² 将局部时间平滑与变化检测结合;与空间导数 gₓ、gᵧ 一起组成时空梯度。把解析核采样为有限数组时,尺度、支撑长度和时空单位共同决定近似质量。对称核用于流式输出时仍须处理未来帧与延迟。

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

      As in the spatial case, it is useful to compute temporal derivatives of spatiotemporal Gaussians: (19.10) where g(x, y, t) is the Gaussian as written in equation (19.7). We can compute the spatiotemporal gradient of a Gaussian: (19.11) We can use the analytic form of the spatiotemporal Gaussian derivatives from equation (19.11) to discretize the filter, by taking samples at discrete locations in space and

    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:摘录中关于时间高斯导数估计变化率的定义、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:As in the spatial case, it is useful to compute temporal derivatives of spatiotemporal Gaussians: (19.10) where g(x, y, t) is the Gaussian as written in equation (19.7). We can compute the spatiotemporal gradient of a Gaussian: (19.11) We can use the analytic form of the spatiotemporal Gaussian derivatives from equation (19.11) to discretize the filter, by taking samples at discrete locations in space and