知识卡片

普通读书笔记卡 · 8831.a · 生成模型:GPT-6

域外成功不等于假设无关

理解等级:中级

域外泛化指系统处理超出设计或训练范围的输入。示例中叠放、硬阴影和互相遮挡违反原假设,却偶然得到合理结果,这不证明假设可任意取消。应明确哪类配置改变、哪些约束受影响,并用更多受控例子判断边界。

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

      simple world. Figure 2.11 shows the result when the input image violates several assumptions we made (shadows are not soft and the green cube occludes the red one) and when it has one object on top of the other, a configuration that we did not consider while designing the algorithm. In this case the result happens to be good, but it could have gone wrong.

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第02章《一个简单视觉系统》(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/c002b.xhtml) - 结论依据:摘录中关于域外成功不等于假设无关的定义、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:simple world. Figure 2.11 shows the result when the input image violates several assumptions we made (shadows are not soft and the green cube occludes the red one) and when it has one object on top of the other, a configuration that we did not consider while designing the algorithm. In this case the result happens to be good, but it could have gone wrong.