Towards Dual-Brain Minimal Sufficient Representation for Vision-Language Navigation

arXiv:2607.23181v1 Announce Type: new Abstract: Vision-and-Language Navigation in continuous environments (VLN-CE) requires an agent to ground language in egocentric observations and plan in unseen scenes. Although recent multimodal large models and world-model-based methods have improved navigation, they often preserve excessive task-irrelevant detail, weakening generalization and increasing computational burden. We propose BrainNav, a navigation framework grounded in the Principle of Minimal Sufficiency. BrainNav consists of three components: a Logical Anchor Model that implements instruction-aware selective perception to suppress environmental noise, a Minimalist Constraint Alignment module that serves as a compact cross-modal bottleneck, efficiently synchronizing discrete linguistic intent with continuous latent dynamics while filtering out redundant information, and a Compression World Model that predicts action-conditioned states within a condensed, low-rank latent space. These modules align semantic intent with spatial perception, enhancing the agent's robustness and efficiency in complex tasks. Experiments show that BrainNav improves over prior SOTA by 2.0 % / 1.0 in SR/SPL on R2R-CE val-unseen and 0.94 % / 0.78 on RxR-CE val-unseen. These results indicate that minimally sufficient world representations provide an effective foundation for robust VLN.

Related

• FutureNav: Unified World-Action Modeling for Vision-and-Language Navigation
• Hierarchical Semantic-Augmented Navigation: Optimal Transport and Graph-Driven Reasoning for Vision-Language Navigation
• SpaceVLN: A Zero-Shot Vision-and-Language Navigation Agent with Online Spatial Cognitive Memory and Reasoning
• From Region Arrival to Instance-Level Grounding in Vision-and-Language Navigation
• Generalizable VLA Finetuning via Representation Anchoring and Language-Action Alignment

Source: arXiv cs.CV | 2026-07-28