Uncertainty-Aware Gaussian Map for Vision-Language Navigation 事件

Uncertainty-Aware Gaussian Map for Vision-Language Navigation arXiv:2605.26503v1 Announce Type: new Abstract: Vision-Language Navigation (VLN) requires an agent to navigate 3D environments following natural language instructions. During navigation, existing agents commonly encounter perceptual uncertainty, such as insufficient evidence for reliable grounding or ambiguity in interpreting spatial cues, yet they typically ignore such information when predicting actions. In this work, we explicitly