课程定位
把感知、规划与控制集成成完整机器人系统。 本页保留官方档案与入口,具体内容与考核以当季官网为准。
主要内容
本页为延伸课程速览,主要内容见下方官方描述。
先修
见官方描述(延伸课程暂未单独整理先修)。
考核与工作量
以当季官网为准(延伸课程暂未单独整理考核)。
适合谁
对「构建 AI 机器人」方向感兴趣的人。
学习建议
先读官方描述和课程官网,再判断是否适合自己。
相关课程
官方课程描述(英文原文)
This course offers a hands-on experiences in building AI-powered robotic quadrupeds. Students will learn the state-of-the-art algorithms and techniques in robot learning and apply them in practice to build a fully functional quadruped. Specifically, the course consists of role-playing paper discussion and hands-on labs. A set of carefully curated papers will be selected at the beginning of the quarter. Each week, everyone reads the same paper, but a small number of student presenters take on specific roles for that paper. The role defines the lens through which the student reads the paper and what they bring to the discussion. Examples of roles include the journalist who summarizes the paper and its contributions based on facts, the paper reviewer who provides critical and insightful criticisms, and the archaeologist who determines where this paper sits in the context of previous and subsequent work. The lab portion of the course focuses on hands-on robot building, programming/training, and evaluation. Students will work in groups of four to develop their quadruped for their sole project in this course. A detailed weekly plan will be proposed and reviewed at the beginning of the quarter. Through weekly labs, students will work toward their final goals of creating an intelligent companion robot dog.Prerequisites: CS123 (or other approved robotics courses), CS107, MATH51