aiqbench analysis
This research introduces scene graphs as a structured memory mechanism to enhance robotic imitation learning, enabling agents to retain historical context and reason over extended time horizons in partially observed environments. Experimental results on simulated and real-world manipulation tasks demonstrate that this approach significantly improves policy performance, particularly for long-term reasoning and generalization. The work addresses critical scalability challenges in deploying autonomous robots within complex, dynamic settings like homes and offices.
Expanding Spatial and Temporal Context for Robotic Imitation Learning With Scene Graphs
arXiv:2606.01072v2 Announce Type: replace Abstract: Imitation learning enables robots to learn how to execute tasks via observation. However, real-world environments like homes and offices are often severely partially observed due to their large spatial scales. In addition, many tasks involve executing a series of subtasks requiring autonomous robots to reason over extended time horizons. To address these challenges, we propose using scene graphs as an explicit and structured memory mechanism in imitation learning. By maintaining a dynamic scene graph that captures object-centric relationships and their evolution over time, our method allows the agent to retain relevant historical context during task execution to efficiently reason over incrementally accrued scene information. Our experiments on simulated mobile manipulation and real-world tabletop manipulation demonstrate that our approach substantially improves policy performance, particularly in settings that demand long-term reasoning and robust generalization under partial observability.
Abstract:Imitation learning enables robots to learn how to execute tasks via observation. However, real-world environments like homes and offices are often severely partially observed due to their large spatial scales. In addition, many tasks involve executing a series of subtasks requiring autonomous robots to reason over extended time horizons. To address these challenges, we propose using scene graphs as an explicit and structured memory mechanism in imitation learning. By maintaining a dynamic scene graph that captures object-centric relationships and their evolution over time, our method allows the agent to retain relevant historical context during task execution to efficiently reason over incrementally accrued scene information. Our experiments on simulated mobile manipulation and real-world tabletop manipulation demonstrate that our approach substantially improves policy performance, particularly in settings that demand long-term reasoning and robust generalization under partial observability.
Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
arXiv:2606.01072 [cs.RO]
(or arXiv:2606.01072v2 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2606.01072 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 232 words
Related coverage
A Hierarchical Spatiotemporal Action Tokenizer for In-Context Imitation Learning in Robotics
arXiv · Sep 17, 2026
Visual Navigation Transformer with Pose Attention
arXiv · Sep 22, 2026
Instant-Fold: In-Context Imitation Learning for Deformable Object Manipulation
arXiv · Jun 4, 2026
R2RDreamer: 3D-aware Data Augmentation for Spatially-generalized 2D Manipulation Policies
arXiv · Jun 16, 2026