← 返回起始笔记位置

知识卡片

普通读书笔记卡 · 8933.c · 生成模型:GPT-6

感受野区分支撑与影响

理解等级:中级

理论感受野是某输出在计算图上可能依赖的输入位置集合;真实权重、激活门控和输入内容会改变各位置实际影响大小。深层局部卷积通常扩大可能依赖范围,但不证明每个远处像素都被有效使用。局部、全局和有效影响应分别讨论。

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

      Receptive fields are another important concept when working with CNNs. In chapter 1, we learned about history of receptive fields in neuroscience. As a reminder, the receptive field of a neuron is the region of the input signal that the neuron is sensitive to, i.e. its support. In multilayer perceptrons (MLPs, chapter 12), the receptive field of each neuron is the entire input vector since

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第24章《卷积神经网络》(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/c024b.xhtml) - 结论依据:摘录中关于感受野区分支撑与影响的定义、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:Receptive fields are another important concept when working with CNNs. In chapter 1, we learned about history of receptive fields in neuroscience. As a reminder, the receptive field of a neuron is the region of the input signal that the neuron is sensitive to, i.e. its support. In multilayer perceptrons (MLPs, chapter 12), the receptive field of each neuron is the entire input vector since