Key Features:
1. It transforms verbal commands from humans into textual instructions, which are then processed by LLMs in the cloud.
2. LLMs generate specific driving codes that are executed by the autonomous vehicle, adjusting driving behaviors and control parameters to align with the human preferences.
3. A memory module in the vehicle stores all human-vehicle interaction data, ensuring each driving experience is personalized based on the human’s historical preferences and commands.

Talk2Drive framework architecture. A human's spoken instructions are processed by cloud-based LLMs, which synthesize contextual data C from weather, traffic conditions, and local traffic rules information. The LLMs generate executable codes P that are communicated to the vehicle's Electronic Control Unit (ECU). These codes operate the actuation of vehicle controls, ensuring that the human's intent is translated into safe and personalized driving actions. A memory module archives every command I, its resultant codes P, and subsequent user feedback F, ensuring continuous refinement of the personalized driving experience.

The initial step in the Talk2Drive framework involves directly receiving arbitrary verbal commands from humans. Utilizing cutting-edge voice recognition technology, specifically the open-source API Whisper, these verbal commands are accurately captured and then translated into textual instructions (I).The LLMs then generate corresponding codes (P) based on this interpretation.The generated codes P are sent back from the cloud to the vehicle's Electronic Control Unit (ECU), where they are executed. The code execution involves adjusting basic driving behaviors and various parameters in the vehicle’s planning and control systems. After ECU executes these codes, the vehicle's actuators control the throttle, brakes, gear selection, and steering to realize the driving behavior specified by the LLM-generated codes through the CAN bus and drive-by-wire system.