Video-Holmes benchmark shows top MLLMs achieve at most 45% accuracy on tasks needing integration of multiple clues from suspense films, unlike existing perception-focused tests.
Activitynet-qa: A dataset for understanding complex web videos via question answering
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
dataset 1
citation-polarity summary
fields
cs.CV 2roles
dataset 1polarities
use dataset 1representative citing papers
LLaVA-OneVision is the first single open LMM to simultaneously achieve strong performance in single-image, multi-image, and video scenarios with cross-scenario transfer capabilities.
citing papers explorer
-
Video-Holmes: Can MLLM Think Like Holmes for Complex Video Reasoning?
Video-Holmes benchmark shows top MLLMs achieve at most 45% accuracy on tasks needing integration of multiple clues from suspense films, unlike existing perception-focused tests.
-
LLaVA-OneVision: Easy Visual Task Transfer
LLaVA-OneVision is the first single open LMM to simultaneously achieve strong performance in single-image, multi-image, and video scenarios with cross-scenario transfer capabilities.