A new dashcam video question-answering benchmark reveals that top vision-language models plateau at 66% accuracy when reasoning about safety-critical driving incidents.
The cityscapes dataset for semantic urban scene understand- ing
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
AUTOPILOT VQA: Benchmarking Vision-Language Models for Incident-Centric Dashcam Understanding
A new dashcam video question-answering benchmark reveals that top vision-language models plateau at 66% accuracy when reasoning about safety-critical driving incidents.