← 返回起始笔记位置

知识卡片

普通读书笔记卡 · 9104.b · 生成模型:GPT-6

实验要逐项支撑论文主张

理解等级:中级

结论需有相应定量或定性证据,比较基线、数据与指标应说明。新任务没有同名基准时,可选相关方法作参照并明确适配差异,不能把为别的任务设计的方法未经合理设置就当作公平弱基线。

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

      In earlier days of computer vision research, good ideas could be presented with only plausibility arguments to support them. However, modern standards require that assertions in a paper be backed up with experimental evidence. Readers will be expecting a table quantitatively comparing the performance of your algorithm with that of the previous state-of-the-art. If you are proposing a new task for which there aren't previous

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第53章《如何写论文》(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/c053.xhtml) - 结论依据:摘录中关于实验要逐项支撑论文主张的定义、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:In earlier days of computer vision research, good ideas could be presented with only plausibility arguments to support them. However, modern standards require that assertions in a paper be backed up with experimental evidence. Readers will be expecting a table quantitatively comparing the performance of your algorithm with that of the previous state-of-the-art. If you are proposing a new task for which there aren't previous