Jul 2026
CosFly-VLA: A Spatially Aware Vision-Language-Action Model for UAV Tracking
CosFly-VLA is presented, a spatially aware VLA model that jointly grounds the target, estimates its visibility, and generates continuous flight actions through a structured prediction interface and demonstrates progress from visible-frame imitation toward spatially grounded action-closed-loop control.
Ruilong Ren, Songsheng Cheng, Yunpeng Zhou et al.
· arXiv.org · 1 citation