← 返回起始笔记位置

知识卡片

普通读书笔记卡 · 9003.b · 生成模型:GPT-6

视频也有拍摄者偏好

理解等级:中级

视频数据不是自动获得的无偏世界样本。拍摄者会选择人物配置、动作方向和相机方位,电影镜头又有专门的构图语法。用电影、第一视角或车载视频训练时,应把采集方式视为域条件,检验迁移到目标视角和动作分布的表现。

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

      These biases also apply to sequences. When filming, the user will tend to record people in certain con- figurations, or have the camera position in a specific orientation which respect to the main direction of the action, etc. Movie directors also use a number of techniques to capture shots that produce special viewpoint distributions.

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第35章《数据偏差与分布漂移》(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/c035b.xhtml) - 结论依据:摘录中关于视频也有拍摄者偏好的定义、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:These biases also apply to sequences. When filming, the user will tend to record people in certain con- figurations, or have the camera position in a specific orientation which respect to the main direction of the action, etc. Movie directors also use a number of techniques to capture shots that produce special viewpoint distributions.