NuRisk is a new VQA dataset for agent-level risk assessment in autonomous driving that benchmarks VLMs at 33% peak accuracy and shows a fine-tuned 7B model reaching 41% with 75% lower latency.
Vision foundation model embedding- based semantic anomaly detection
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
representative citing papers
An embedding-based method detects and localizes anomalies in driving scenes via nearest-neighbor similarity to a single reference image, evaluated on benchmarks and real automated vehicles.
citing papers explorer
-
NuRisk: A Visual Question Answering Dataset for Agent-Level Risk Assessment in Autonomous Driving
NuRisk is a new VQA dataset for agent-level risk assessment in autonomous driving that benchmarks VLMs at 33% peak accuracy and shows a fine-tuned 7B model reaching 41% with 75% lower latency.
-
Real-World On-Vehicle Evaluation of Embedding-Based Anomaly Detection
An embedding-based method detects and localizes anomalies in driving scenes via nearest-neighbor similarity to a single reference image, evaluated on benchmarks and real automated vehicles.