Fu, T.; Zhao, M.; Niu, K.; Peng, K.; Li, B. OmniPT: Unleashing the Potential of Large Vision Language Models for Pedestrian Tracking and Understanding. AAAI 2026, 40, 4031-4039.