Sp guide improves drone navigation using spatial knowledge

SP-Guide: Enhancing Aerial Vision-Language Navigation with Spatial Priors

Computer Vision and Pattern RecognitionArtificial Intelligence

Summary

Finding a location described by language from a drone’s viewpoint in large outdoor spaces is tough. The authors propose SP-Guide, a system that better matches what the drone sees with instructions by using spatial knowledge as a guide. This method helps the drone understand where to go more accurately, especially in places it hasn’t flown before. Tests show that using this spatial prior helps the drone reach targets more successfully.

What this means in practice

  • For drone navigation developers: Develop UAV systems that reach language-described outdoor targets more reliably using spatial prior embeddings for better alignment of views and instructions.
  • For robotics platform engineers: Enhance outdoor robot navigation by integrating spatially grounded priors to improve goal prediction and terminal accuracy in complex environments.

Authors

Haoxiang Tang, Dong Zhou, Shen Tan, Minxuan Jiang, Xitong Li

Abstract

Aerial vision-language navigation (AVLN) requires a UAV to reach a language-described target from partial egocentric observations while reasoning over large outdoor scenes. Existing approaches typically fuse view-centered RGB observations and globally anchored history maps as generic visual embeddings, leaving cross-modal spatial alignment to be learned from sparse navigation supervision. We present SP-Guide, a transformer- based goal-prediction framework that addresses this weakness of alignment through a Spatial Prior Cross-Modal Encoder, an Instruction Enhanced Decoder, and a Goal Convergence Filter for terminal target refinement. Experiments implemented on the CityNav benchmark show that SP-Guide improves success rate over the representative baselines, with the largest gains on unseen environments. These results indicate that physically grounded spatial priors provide an effective inductive bias for robust AVLN.