Visual-Tactile Multimodality for Following Deformable Linear Objects Using Reinforcement Learning