TY  - RPRT
TI  - Can Vision-Language Models Assess Proxemic Risk from Egocentric Robot Images?
AU  - Vladyslava Rudas
AU  - Dmytro Kuzmenko
PY  - 2026
UR  - https://arxiv.org/abs/2608.12515
ID  - 2608.12515
ER  -