← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 8921.c · 生成模型:GPT-6

速度选择由频率平面聚合

理解等级:中级

可先用时空Gabor正交对得到对相位较不敏感的频率响应,再按候选速度对应的频谱平面组合多路能量,形成速度选择单元。最强候选给出模型内的速度估计,但重复纹理、低纹理、遮挡与多运动仍可产生歧义;候选离散化也限制精度。

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

      We can use quadrature pairs of oriented filters in space-time to find motion speed and direction in the video signal. We just need to find the space-time orientation of strongest response. Figure 22.16(a) shows a set of Gabor filters sampling the space-time frequency domain. When the input contains a moving signal, we can use a set of filters to identify the plane in the Fourier

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第22章《滤波器组》(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/c022.xhtml) - 结论依据:摘录中关于速度选择由频率平面聚合的定义、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:We can use quadrature pairs of oriented filters in space-time to find motion speed and direction in the video signal. We just need to find the space-time orientation of strongest response. Figure 22.16(a) shows a set of Gabor filters sampling the space-time frequency domain. When the input contains a moving signal, we can use a set of filters to identify the plane in the Fourier