Embodied AI

AIAA 4220, Fall 2026

Instructors: Junwei Liang


Home Schedule

Lecture:

Date and Time: 09:00AM - 11:50AM, Monday
Location: Rm 233, E1
Websites: Instructure
Zoom: 628-334-1826 PW: 12345 (In-person expected) (Recordings will be available after each class)

Course Description:

This course aims to provide students with key principles and algorithms to build modern Embodied AI systems. Key topics include machine perception, embodied navigation and manipulation algorithms. Through this course, students will learn and practice the foundational principles, techniques, and tools to build new Embodied AI systems.

Target Audience/ Prerequisites: This is a course primarily for junior and senior students.

Course Work:
Grading      Grading will be based on best 5 of 6 in-class quizzes (25%), one paper presentation (25%), one machine perception homework (25%) and one simulation project (25%).
In-Class Quizzes
  • The in-class quizzes are designed to complement the lectures and making sure you understand the key knowledge about the course.
  • There are 6 in-class quizzes over the semester. We will take your best 5 out of the 6 quizzs as your grade.
Paper Presentation
  • You will select 1 paper from our list of Embodied AI research to present. This should show that you understand the paper, and you could present it to students who do not know the paper well.
  • This is a individual assignment.
  • Project 1 is worth 25% credit. You will be graded based on Understanding & Content (40%), Presentation & Communication (30%), Critical Analysis & Discussion (30%), by the teacher and TAs. Grading rubrics.
Machine Perception Homework
  • You will train an ego-vision object detection model and submit to Kaggle leaderboard.
  • This is a individual assignment.
  • Instructions are here.
  • Grading will be based on the leaderboard scores and a written report.
Simulation Project
  • You will select one simulation project from a project pool to work on. Reproduce the baselines and add 1-2 improvements you choose.
  • Can be done in groups of 2-3 people.

Instructors:

Junwei Liang
Rm 304, E4
Office hours: Wednesday 10:00AM - 11:30AM

Teaching Assistants:

Yuxin He , yhe469@connect.hkust-gz.edu.cn
Jinghao Huang , jhuang401@connect.hkust-gz.edu.cn
Compute resource: Courtesy of AI thrust, we can provide at least ¥200 for each student to use the university HPC. Ask the TA for account and password.
Collaboration among Students: We encourage collaboration between students and studying materials in groups when the purpose of this is to facilitate learning, not to circumvent problems. It is allowed to seek help from other students in understanding the material needed to solve a particular problem. However, students must submit individual material and solutions, unless otherwise specified. Students should declare any collaboration on the first page of homework assignments (or equivalently on exercises). If the instructors believe the collaboration is improper, your grade may be affected.

Use of LLM/VLMs (ChatGPT, Gemini, etc.): We encourage you to use ChatGPT to solve the homework assignment problems (because likely in the future everyone will use it for coding). However, you will need to declare using ChatGPT/LLMs in the homework README, as well as the prompts you use. But you are not allowed to use LLM for in-class test!!!