← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 9225 · 生成模型:GPT-6

感知支架要保留集成边界

理解等级:中级

可用仿真自带检测分割或现成网络,让学生聚焦运动学,但应明确被抽象掉的感知误差。简单箱内抓取与避障放置需要的规划能力不同,任务升级时不能默认为旧支架仍足够。

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

      Object recognition and segmentation are good motivation for introducing convolutional neural networks (chapter 10) as well as appropriate open-source tools that the students can use as a black box. Simulators such as Webots also provide object detection and segmentation, allowing the instructor to focus only on the kinematic aspects of autonomy.

    参考来源

    - 位置:《Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)》附录F《附录F:课程方案》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Introduction to Autonomous Robots Mechanisms, Sensors, Actuators, and Algorithms (Nikolaus Correll, Bradley Hayes etc.)/_epub-src/OEBPS/xhtml/appendixF.xhtml) - 结论依据:摘录中关于感知支架要保留集成边界的定义、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:Object recognition and segmentation are good motivation for introducing convolutional neural networks (chapter 10) as well as appropriate open-source tools that the students can use as a black box. Simulators such as Webots also provide object detection and segmentation, allowing the instructor to focus only on the kinematic aspects of autonomy.