CARLA

CARLA is a large-scale dataset and benchmark that contains over 890K frames, 176 tasks, and 85 diverse objects, used to evaluate multimodal tracking in dynamic urban environments for Unmanned Aerial Vehicles (UAVs).

Papers using CARLA (8)

UAV-Track VLA: Embodied Aerial Tracking via Vision-Language-Action Models2026Comparative Analysis of Patch Attack on VLM-Based Autonomous Driving Architectures2026AutoFocus-IL: VLM-based Saliency Maps for Data-Efficient Visual Imitation Learning without Extra Human Annotations2025AutoVLA: A Vision-Language-Action Model for End-to-End Autonomous Driving with Adaptive Reasoning and Reinforcement Fine-Tuning2025Exploring Multimodal Implicit Behavior Learning For Vehicle Navigation In Simulated Cities2025X-driver: Explainable Autonomous Driving With Vision-language Models2025RLM: A Vision-language Model Approach For Radar Scene Understanding2025Efficient Policy Adaptation with Contrastive Prompt Ensemble for Embodied Agents2024 · 3 cites