Search papers, labs, and topics across Lattice.
SAP-Nav introduces a novel framework for hierarchical open-vocabulary object navigation (OVON) that integrates active perception to enhance spatial grounding and target verification in unseen environments. By constructing a Queryable Spatial-Semantic Representation from actively acquired views, the system allows for dynamic spatial queries and optimizes viewpoint selection to ensure effective target verification. Experimental results demonstrate a significant 12.2% improvement in success rate over existing training-based methods, showcasing its robustness in both simulated and real-world scenarios.
Achieving a 12.2% boost in navigation success rates, SAP-Nav redefines how agents can dynamically interact with and understand their environments without prior training.
Hierarchical open-vocabulary object navigation (OVON) requires agents to follow free-form instructions that may specify targets through scene-, room-, region-, and instance-level cues in unseen environments. Although recent work LangMap has formalized this setting, reliably solving it under partial observations remains challenging: spatial grounding requires persistent environment-level evidence, whereas target verification requires clear and discriminative candidate views. We present SAP-Nav, a fully online, zero-shot framework that addresses both requirements through active perception. SAP-Nav incrementally constructs a Queryable Spatial-Semantic Representation from actively acquired room views, enabling spatial semantic queries from any explored location. It further employs Active Viewpoint Verification to assess whether the current observation provides sufficient evidence and, when necessary, reposition the agent to a more informative viewpoint before verifying candidates against category and attribute constraints. Although designed for hierarchical OVON, SAP-Nav supports both hierarchical and standard category-level OVON without task-specific training or precomputed scene maps. Experiments on LangMap and HM3D-OVON show that SAP-Nav achieves the overall best performance, including a 12.2% improvement in SR over training-based methods on region-level navigation. Real-world robot experiments further demonstrate its practical feasibility. Code will be made publicly available upon acceptance.
