TAR and TAR-Bench provide a ten-task traffic anomaly reasoning dataset and benchmark, and fine-tuning on the multi-task chain-of-thought data raises VLM mean scores by about 21 points.
CARLA: An open urban driving simulator
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2026 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
From Detection to Understanding: TAR and TAR-Bench for Multi-Task Traffic Anomaly Reasoning
TAR and TAR-Bench provide a ten-task traffic anomaly reasoning dataset and benchmark, and fine-tuning on the multi-task chain-of-thought data raises VLM mean scores by about 21 points.