HoloCount is a three-tier visual counting benchmark showing that MLLMs fail systematically on analytical reasoning, high-density scenes, and linguistic prior conflicts, with even the best models dropping below 50% accuracy on dense counting.
Ec-bench: Enumeration and counting benchmark for ultra-long videos, 2026
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
HoloCount: A Holistic Visual Counting Benchmark for MLLMs
HoloCount is a three-tier visual counting benchmark showing that MLLMs fail systematically on analytical reasoning, high-density scenes, and linguistic prior conflicts, with even the best models dropping below 50% accuracy on dense counting.