TY - RPRT TI - Learning to Navigate Under Imperfect Perception: Conformalised Segmentation for Safe Reinforcement Learning AU - Daniel Bethell AU - Simos Gerasimou AU - Radu Calinescu AU - Calum Imrie PY - 2025 UR - https://arxiv.org/abs/2510.18485 ID - 2510.18485 ER -