CS269

CS269: AI Applications in Robotics (fall 2024)

Schedule

Date	Topic	Readings	Deadline
Week 1 Thu, Sep 26	Introduction & Robotics Basics	Please checkout our Course Policies. Slides	Sign up for 2 presenter slots by Tuesday Oct 1st.
Week 2 Tue, Oct 1	Robot Learning Overview	Building Machines That Learn and Think Like People A Review of Robot Learning for Manipulation: Challenges, Representations, and Algorithms Five Breakthroughs: A First Approximation of Brain Evolution From Early Bilaterians to Humans	Paper response due the day before class at 9PM.
Week 2 Thu, Oct 3	Robot Perception	Segment Anything Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data Vision-based Manipulation from Single Human Video with Open-World Object Graphs	Paper response due the day before class at 9PM.
Week 3 Tue, Oct 8	Transformers	Behavior Transformers: Cloning k modes with one stone Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation Decision Transformer: Reinforcement Learning via Sequence Modeling	Paper response due the day before class at 9PM.
Week 3 Thu, Oct 10	Diffusion Models	Planning with Diffusion for Flexible Behavior Synthesis Diffusion Policy: Visuomotor Policy Learning via Action Diffusion Parallel Sampling of Diffusion Models	Paper response due the day before class at 9PM.
Week 4 Tue, Oct 15	Data Quality	What Matters in Learning from Offline Human Demonstrations for Robot Manipulation Data Quality in Imitation Learning Re-Mix: Optimizing Data Mixtures for Large Scale Imitation Learning	Paper response due the day before class at 9PM.
Week 4 Thu, Oct 17	Data Curation	ThriftyDAgger: Budget-Aware Novelty and Risk Gating for Interactive Imitation Learning Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real World Using Simulation and Domain Adaptation to Improve Efficiency of Deep Robotic Grasping	Paper response due the day before class at 9PM. Project Proposal due Oct 18 at 11:59PM
Week 5 Tue, Oct 22	Large Language Models	SayCan: Grounding Language in Robotic Affordances Code as Policies: Language Model Programs for Embodied Control Robots That Ask For Help: Uncertainty Alignment for Large Language Model Planners	Paper response due the day before class at 9PM.
Week 5 Thu, Oct 24	Large Robotics Models	RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control Manipulate-Anything: Automating Real-World Robots using Vision-Language Models OpenVLA: An Open-Source Vision-Language-Action Model	Paper response due the day before class at 9PM.
Week 6 Tue, Oct 29	Reinforcement Learning	Proximal Policy Optimization Algorithms Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor Imitation Bootstrapped Reinforcement Learning	Paper response due the day before class at 9PM.
Week 6 Thu, Oct 31	Inverse Reinforcement Learning	Apprenticeship Learning via Inverse Reinforcement Learning Generative Adversarial Imitation Learning Learning from Suboptimal Demonstration via Self-Supervised Reward Regression	Paper response due the day before class at 9PM.
Week 7 Tue, Nov 5	Representation Learning	Language-Driven Representation Learning for Robotics HRP: Human Affordances for Robotic Pre-Training UniSim: Learning Interactive Real-World Simulators	Paper response due the day before class at 9PM. Project Milestone Review due Nov 8 at 11:59PM
Week 7 Thu, Nov 7	Meta Learning	Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks Reptile: A Scalable Meta-Learning Algorithm Octo: An Open-Source Generalist Robot Policy	Paper response due the day before class at 9PM.
Week 8 Tue, Nov 12	Human-Robot Interaction I	Legibility and Predictability of Robot Motion Learning Latent Representations to Influence Multi-Agent Interaction Feel the Bite: Robot-Assisted Inside-Mouth Bite Transfer using Robust Mouth Perception and Physical Interaction-Aware Control	Paper response due the day before class at 9PM.
Week 8 Thu, Nov 14	Human-Robot Interaction II	Shared Autonomy via Deep Reinforcement Learning SIRL: Similarity-based Implicit Representation Learning Is More Autonomy Always Better? Exploring Preferences of Users with Mobility Impairments in Robot-Assisted Feeding	Paper response due the day before class at 9PM.
Week 9 Tue, Nov 19	Emerging Topics	KITE: Keypoint-Conditioned Policies for Semantic Manipulation FlowRetrieval: Flow-Guided Data Retrieval for Few-Shot Imitation Learning LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learning	Paper response due the day before class at 9PM.
Week 9 Thu, Nov 21	Lecture
Week 10 Tue, Nov 26	Office Hour for Projects
Week 10 Thu, Nov 28	Thanksgiving Day (No class)
Week 11 Thu, Dec 3	Final Project Presentation
Week 11 Thu, Dec 5	Final Project Presentation		Project report due Dec 5 at 11:59PM (firm).

Additional resources:

Part of this course is based on: