知识卡片

普通读书笔记卡 · 9082.d · 生成模型:GPT-6

多帧信息需保留时间可见性

理解等级:中级

增加输入帧可提供更长轨迹和额外遮挡线索,但也带来更复杂对应、延迟与计算成本。未来帧在离线任务可用,在线任务可能不可用;不能把多帧模型的结果无条件放到实时双帧设定比较。

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

      The network can also take as input multiple frames and not just two. 49.3 Concluding Remarks Supervised and unsupervised learning-based methods are now the state-of-the-art in motion estimation. But an accurate solution is still missing. One important question is, do we really need learning in order to solve this problem? Should we abandon the derivation of physically motivated algorithms for motion estimation that require no

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第49章《学习运动估计》(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/c049.xhtml) - 结论依据:摘录中关于多帧信息需保留时间可见性的定义、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:The network can also take as input multiple frames and not just two. 49.3 Concluding Remarks Supervised and unsupervised learning-based methods are now the state-of-the-art in motion estimation. But an accurate solution is still missing. One important question is, do we really need learning in order to solve this problem? Should we abandon the derivation of physically motivated algorithms for motion estimation that require no