Text this: Vision-Language Model-Driven Human-Vehicle Interaction for Autonomous Driving: Status, Challenge, and Innovation