TAR and TAR-Bench serve as the official training and in-domain evaluation resources for AI City Challenge 2026 Track 3 and highlight persistent limitations in temporal precision and causal attribution.
Han Zhang, Yi-Lin Zhao, Zaid Pervaiz Bhat et al.· 1 citation
VANTAGE-Bench is introduced, a benchmark measuring the "Infrastructure AI Gap", which spans three operational domains, unifies image and video evaluation across semantic, spatial, temporal, and spatio-temporal capabilities, and moves beyond multiple-choice to eight task formulations including dense captioning and spati...