그래프 기반 상태 표현을 활용한 작업 계획 알고리즘 개발
Task Planning Algorithm with Graph-based State Representation
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
The ability to understand given environments and plan a sequence of actions leading to goal state is crucial for personal service robots. With recent advancements in deep learning, numerous studies have proposed methods for state representation in planning. However, previous works lack explicit information about relationships between objects when the state observation is converted to a single visual embedding containing all state information. In this paper, we introduce graph-based state representation that incorporates both object and relationship features. To leverage these advantages in addressing the task planning problem, we propose a Graph Neural Network (GNN)-based subgoal prediction model. This model can extract rich information about object and their interconnected relationships from given state graph. Moreover, a search-based algorithm is integrated with pre-trained subgoal prediction model and state transition module to explore diverse states and find proper sequence of subgoals. The proposed method is trained with synthetic task dataset collected in simulation environment, demonstrating a higher success rate with fewer additional searches compared to baseline methods.
Abstract
The ability to understand given environments and plan a sequence of actions leading to goal state is crucial for personal service robots. With recent advancements in deep learning, numerous studies have proposed methods for state representation in planning. However, previous works lack explicit information about relationships between objects when the state observation is converted to a single visual embedding containing all state information. In this paper, we introduce graph-based state representation that incorporates both object and relationship features. To leverage these advantages in addressing the task planning problem, we propose a Graph Neural Network (GNN)-based subgoal prediction model. This model can extract rich information about object and their interconnected relationships from given state graph. Moreover, a search-based algorithm is integrated with pre-trained subgoal prediction model and state transition module to explore diverse states and find proper sequence of subgoals. The proposed method is trained with synthetic task dataset collected in simulation environment, demonstrating a higher success rate with fewer additional searches compared to baseline methods.
참고문헌 (9)
- M. D. Graaf, "Why do they erfuse to use my robot?: Reasons for non-use derived from a long-term home study", 2017 ACM, 2017, 224~233
- H. Gou, "Recent Trends ni Task and Motion Planning for Robotics: A Survey", ACM Comput. Surv, 55(13), pp.1~36 (2023)[DOI]
- PDDL — The Planning Domain Definition Language, "M. Ghallab", Yale Center for Computational Vision and Control, ~ (1998)
- C. Paxton, "Visual robot task planning", 2019 International conference on robotics and automation(ICRA), pp.8832~8838 (2019)
- T. Silver, "PDDL planning with pretrained large language models", NeurIPS 2022 foundation models for decision making workshop, pp.2022~ (2022)
- J. Zhou, "Graph neural networks : A review of methods and applications", AI open, 1, pp.57~81 (2020)[DOI]
- T. Silver, "Planning with learned object importance in large problem instances using graph neural networks", Proceedings of the AAAI conference on artificial intelligence, pp.11962~11971 (2021)
- K. He, "Deep ersidual elarning for image recognition", Proceedings of the IEEE conference on computer vision and pattern recognition, pp.770~778 (2016)
- J. J. Kuffner, "RRT-connect : An efficient approach to single-query path planning", Proceedings 2000ICRA. Millennium Conference. IEEE International Conference on Robotics and Automation. Symposia Proceedings(Cat. No. 00CH37065), 2, pp.995~1001 (2000)