{"ID":2863692,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2509.24910","arxiv_id":"2509.24910","title":"Learning Goal-Oriented Vision-and-Language Navigation with Self-Improving Demonstrations at Scale","abstract":"Goal-oriented vision-language navigation requires robust exploration capabilities for agents to navigate to specified goals in unknown environments without step-by-step instructions. Existing methods tend to exclusively utilize shortest-path trajectories, lacking effective exploration priors for training navigation agents. To address the above challenges, we present SID, a goal-oriented vision-and-language navigation learning approach with Self-Improving Demonstrations. Specifically, SID learns an initial agent on the shortest-path data sampled from environments and then leverages this agent to generate novel exploration trajectories. The novel rollouts provide demonstrations with stronger exploration strategies to train a better agent, which in turn produces higher-quality agent demonstrations for the next round of training. We show that this iterative self-improving pipeline readily scales to new environments, and the resulting demonstrations are highly transferable, elevating the performance ceiling across a variety of vision-and-language navigation tasks. Extensive experiments demonstrate that SID significantly boosts the exploration capabilities and generalization of navigation agents. The resulting agent achieves new state-of-the-art performance on goal-oriented vision-and-language navigation benchmarks, including REVERIE, SOON as well as strong transferability to object-goal navigation and VLN-CE. It notably achieves a 50.9% success rate on the unseen validation splits of SOON, surpassing prior leading approaches by a margin of 13.9%.","short_abstract":"Goal-oriented vision-language navigation requires robust exploration capabilities for agents to navigate to specified goals in unknown environments without step-by-step instructions. Existing methods tend to exclusively utilize shortest-path trajectories, lacking effective exploration priors for training navigation age...","url_abs":"https://arxiv.org/abs/2509.24910","url_pdf":"https://arxiv.org/pdf/2509.24910v2","authors":"[\"Songze Li\",\"Zun Wang\",\"Gengze Zhou\",\"Jialu Li\",\"Xiangyu Zeng\",\"Ziyang Gong\",\"Limin Wang\",\"Yu Qiao\",\"Qi Wu\",\"Mohit Bansal\",\"Yi Wang\"]","published":"2025-09-29T15:15:54Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[\"LoRA\"]","has_code":false}