SS-MAIL: Self-Supervised Multi-Agent Imitation Learning