← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 9078.b · 生成模型:GPT-6

慢速先验与平滑先验不同

理解等级:中级

惩罚Σ(u²+v²)偏好较小速度,惩罚Σ(|∇u|²+|∇v|²)偏好邻近速度相近。前者可压低整体大运动,后者允许恒定大流却容易模糊边界;二者不能同称为限制流快速空间变化的同一种正则。

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

      There are several popular regularization terms. One penalizes large velocities (slow prior): (48.14) Another penalizes variations on the optical flow (smooth prior): (48.15) The photometric loss plays an important role in unsupervised learning-based methods for optical flow estimation as we will discuss later. 48.3.4 Layer-Based Motion Estimation Until now we have not made use of any of the properties of the motion field derived from

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第48章《光流估计》(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/c048.xhtml) - 结论依据:摘录中关于慢速先验与平滑先验不同的定义、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:There are several popular regularization terms. One penalizes large velocities (slow prior): (48.14) Another penalizes variations on the optical flow (smooth prior): (48.15) The photometric loss plays an important role in unsupervised learning-based methods for optical flow estimation as we will discuss later. 48.3.4 Layer-Based Motion Estimation Until now we have not made use of any of the properties of the motion field derived from