Hierarchical Policy Learning

Hierarchical policy learning aims to solve complex tasks by breaking them down into simpler sub-tasks, managed by a hierarchy of policies. Current research focuses on improving subgoal reachability through bidirectional information sharing between policy levels, employing various architectures like options frameworks and model predictive control, and addressing challenges such as hallucination in vision-language models and efficient learning from limited data in offline settings. This approach enhances the robustness, efficiency, and generalizability of reinforcement learning agents, with significant implications for robotics, autonomous systems, and other domains requiring complex sequential decision-making.

Papers

January 9, 2025

From Simple to Complex Skills: The Case of In-Hand Object Reorientation
Haozhi Qi, Brent Yi, Mike Lambeta, Yi Ma, Roberto Calandra, Jitendra Malik
Case Relevance Dexterous Manipulation Diverse Skill Hierarchical Policy Learning Hand Object Reorientation Skill Policy

December 19, 2024

November 6, 2024

Object-Centric Dexterous Manipulation from Human Motion Data
Yuanpei Chen, Chen Wang, Yaodong Yang, C. Karen Liu
Human Motion Robotic Hand Dexterous Manipulation Bimanual Robot Hierarchical Policy Learning

September 30, 2024

HELPD: Mitigating Hallucination of LVLMs by Hierarchical Feedback Learning with Vision-enhanced Penalty Decoding
Fan Yuan, Chi Qin, Xiaogang Xu, Piji Li
Large Vision Language Model Content Hallucination Hierarchical Learning Constrained Decoding Multimodal Hallucination Hallucination Benchmark Hierarchical Policy Learning Gradient Penalty

June 26, 2024

Bidirectional-Reachable Hierarchical Reinforcement Learning with Mutually Responsive Policies
Yu Luo, Fuchun Sun, Tianying Ji, Xianyuan Zhan
Hierarchical Reinforcement Learning Diverse Policy Hierarchical Policy Learning Goal Reachability Exploration Characteristic

March 16, 2024

Reinforcement Learning with Options and State Representation
Ayoub Ghriss, Masashi Sugiyama, Alessandro Lazaric
Reinforcement Learning Markov Decision Process Hierarchical Reinforcement Learning Complex Environment Continuous Option Hierarchical Policy Learning

December 17, 2023

GO-DICE: Goal-Conditioned Option-Aware Offline Imitation Learning via Stationary Distribution Correction Estimation
Abhinav Jain, Vaibhav Unhelkar
Long Term Policy Learning DIstribution Estimation Offline Imitation Offline Imitation Learning Go Game Hierarchical Policy Learning

December 5, 2023

Hierarchical Visual Policy Learning for Long-Horizon Robot Manipulation in Densely Cluttered Scenes
Hecheng Wang, Lizhe Qi, Bin Fang, Yunquan Sun
Hierarchical Reinforcement Learning Cluttered Environment Long Horizon Manipulation Task Cluttered Scene Hierarchical Policy Learning

August 18, 2023

Intrinsically Motivated Hierarchical Policy Learning in Multi-objective Markov Decision Processes
Sherif Abdelfattah, Kathryn Merrick, Jiankun Hu
Reinforcement Learning Multi Objective Reinforcement Learning Hierarchical Policy Learning Multi Objective Markov Decision Process Evolutionary Policy Search

May 22, 2023

Multi-task Hierarchical Adversarial Inverse Reinforcement Learning
Jiayu Chen, Dipesh Tamboli, Tian Lan, Vaneet Aggarwal
Task Specific Policy Adversarial Inverse Reinforcement Learning Hierarchical Policy Learning

November 22, 2022

Evaluation of MPC-based Imitation Learning for Human-like Autonomous Driving
Flavia Sofia Acerbo, Jan Swevers, Tinne Tuytelaars, Tong Duy Son
Global Evaluation Predictive Control Policy Gradient Closed Loop Human Like Autonomous Driving Hierarchical Policy Learning

February 28, 2022

Hierarchical Policy Learning for Mechanical Search
Oussama Zenkri, Ngo Anh Vien, Gerhard Neumann
Deep Reinforcement Learning Action Selection Hierarchical Policy Learning Hierarchical MDP Mechanical Search

February 17, 2022

Multi-Modal Fusion in Contact-Rich Precise Tasks via Hierarchical Policy Learning
Piaopiao Jin, Yinjie Lin, Yanchao Tan, Tiefeng Li, Wei Yang
Sensor Fusion Multi Modal Fusion Contact Rich Visual Feedback Multi Modality Fusion Hierarchical Policy Learning Contact Rich Task