MMDVS-LF: A Multi-Modal Dynamic-Vision-Sensor Line Following Dataset