arxivcs.CV2026-06-26
HumanMoveVQA: Can Video MLLMs reason about human movement in videos?
Pulkit Gera, Faegheh Sardari, Asmar Nadeem, Valentina Bono, Padraig Boulton, Adrian Hilton, et al.
Despite the rapid advance of Multimodal Large Language Models (MLLMs) in high-level video understanding, a fundamental bottleneck remains: these models collapse complex human motion into coarse semantic labels. Existing benchmarks mostly focus on scene-centric events or local joi…