analiza aiqbench
MatchingPolicy advances robotic control by decoupling semantic object matching from action generation, enabling robust few-shot generalization across unseen objects using vision foundation models. This architectural innovation directly addresses limitations in current in-context imitation learning where performance degrades on novel scenarios without requiring extensive retraining. By establishing reliable dense correspondences dynamically, the framework significantly improves out-of-distribution transfer capabilities for autonomous manipulation tasks.
MatchingPolicy: Correspondence-Aware Policy Enables Cross-Object In-Context Learning
arXiv:2608.16715v1 Announce Type: new Abstract: In-context imitation learning enables few-shot policy generalization but struggles to maintain performance on unseen objects and novel scenarios. To address this, we introduce MatchingPolicy, a correspondence-driven framework that explicitly decouples demonstration-to-scene matching from policy learning. Central to our method is a correspondence-aware diffusion policy that conditions robotic actions directly on dense semantic correspondences. This architectural separation resolves the inherent conflict between correspondence identification and action adaptation, enabling robust out-of-distribution transfer. Our framework integrates vision foundation models with a novel two-stage matching algorithm to dynamically establish reliable correspondences. Extensive evaluations on RLBench and real-world manipulation tasks confirm that MatchingPolicy achieves superior few-shot performance, generalizing reliably across unseen object instances and semantic categories.
Abstract:In-context imitation learning enables few-shot policy generalization but struggles to maintain performance on unseen objects and novel scenarios. To address this, we introduce MatchingPolicy, a correspondence-driven framework that explicitly decouples demonstration-to-scene matching from policy learning. Central to our method is a correspondence-aware diffusion policy that conditions robotic actions directly on dense semantic correspondences. This architectural separation resolves the inherent conflict between correspondence identification and action adaptation, enabling robust out-of-distribution transfer. Our framework integrates vision foundation models with a novel two-stage matching algorithm to dynamically establish reliable correspondences. Extensive evaluations on RLBench and real-world manipulation tasks confirm that MatchingPolicy achieves superior few-shot performance, generalizing reliably across unseen object instances and semantic categories.
Robotics (cs.RO)
arXiv:2608.16715 [cs.RO]
(or arXiv:2608.16715v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2608.16715 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
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.
Text integral sub licența deschisă a publicației · 204 cuvinte
Subiecte conexe
Improving CoT Monitorability of Evaluation Awareness via Verbalization Training
LessWrong · 2 oct. 2026
Assessing Information Ecosystems : How Governments Can Get Ahead of Hybrid Threats
carnegieendowment.org · 5 aug. 2026
Links #3: 2026/06 Part 1
LessWrong · 15 iun. 2026
[AINews] Reality Checks on AI News (Yegge shuts down Gas Town, Databricks’ +60% Astra cost)
Latent Space · 17 sept. 2026