{"ID":2873761,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2509.06031","arxiv_id":"2509.06031","title":"GELATO: Multi-Instruction Trajectory Reshaping via Geometry-Aware Multiagent-based Orchestration","abstract":"We present GELATO -- the first language-driven trajectory reshaping framework to embed geometric environment awareness and multi-agent feedback orchestration to support multi-instruction in human-robot interaction scenarios. Unlike prior learning-based methods, our approach automatically registers scene objects as 6D geometric primitives via a VLM-assisted multi-view pipeline, and an LLM translates free-form multiple instructions into explicit, verifiable geometric constraints. These are integrated into a geometric-aware vector field optimization to adapt initial trajectories while preserving smoothness, feasibility, and clearance. We further introduce a multi-agent orchestration with observer-based refinement to handle multi-instruction inputs and interactions among objectives -- increasing success rate without retraining. Simulation and real-world experiments demonstrate our method achieves smoother, safer, and more interpretable trajectory modifications compared to state-of-the-art baselines.","short_abstract":"We present GELATO -- the first language-driven trajectory reshaping framework to embed geometric environment awareness and multi-agent feedback orchestration to support multi-instruction in human-robot interaction scenarios. Unlike prior learning-based methods, our approach automatically registers scene objects as 6D g...","url_abs":"https://arxiv.org/abs/2509.06031","url_pdf":"https://arxiv.org/pdf/2509.06031v2","authors":"[\"Junhui Huang\",\"Yuhe Gong\",\"Changsheng Li\",\"Xingguang Duan\",\"Luis Figueredo\"]","published":"2025-09-07T12:13:51Z","proceeding":"cs.RO","tasks":"[\"cs.RO\"]","methods":"[\"Large Language Model\"]","has_code":false}