AI RESEARCH
EvoMemNav: Efficient Self-Evolving Fine-Grained Memory for Zero-Shot Embodied Navigation
arXiv CS.CV
•
ArXi:2606.03509v1 Announce Type: new Building memory is essential for long-horizon planning in zero-shot embodied navigation. Detector-centric scene graphs often compress observations into sparse nodes, discarding fine-grained visual evidence and accumulating noise, while 3D reconstruction-based methods remain computationally prohibitive. We present EvoMemNa, an efficient, self-evolving, fine-grained memory framework for zero-shot embodied navigation.