Bilateral Deep Reinforcement Learning Approach for Better-than-human Car-following | IEEE Conference Publication | IEEE Xplore