← 返回起始笔记位置

知识卡片

普通读书笔记卡 · 8867.d · 生成模型:GPT-6

图像输出的范围不等于编码

理解等级:中级

本章示例以255·sigmoid(z)把浮点输出限定在8位RGB常用范围内,再按需要量化为整数。它不是所有图像生成的唯一输出形式,也不自动规定颜色空间、传递函数或文件格式。模型目标、存储编码与显示解释要分别确定。

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

      label, and the most common output operation is the softmax function, which we have already encountered in previous chapters. In image synthesis problems, the desired output is typically a 3D array with dimensions N × M × 3, and values in the range [0, 255]. A sigmoid multiplied by 255 is a typical output transformation for this setting. The equations for these two layers are:

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第12章《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/c012.xhtml) - 结论依据:摘录中关于图像输出的范围不等于编码的定义、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:label, and the most common output operation is the softmax function, which we have already encountered in previous chapters. In image synthesis problems, the desired output is typically a 3D array with dimensions N × M × 3, and values in the range [0, 255]. A sigmoid multiplied by 255 is a typical output transformation for this setting. The equations for these two layers are: