DVBench introduces 10,000 expert-annotated questions on crash and near-crash driving videos and reports that no tested vision LLM exceeds 40 percent accuracy under its strict GroupEval scoring.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Are Vision LLMs Road-Ready? A Comprehensive Benchmark for Safety-Critical Driving Video Understanding
DVBench introduces 10,000 expert-annotated questions on crash and near-crash driving videos and reports that no tested vision LLM exceeds 40 percent accuracy under its strict GroupEval scoring.