Dataset for "From Segments to Scenes: Temporal Understanding for Agentic Autonomous Driving via Vision-Language Models"

You can download the benchmark using the following Python code snippet. from datasets import load_dataset

action_duration = load_dataset("vbdai/TAD", name="action_duration") exact_answer_action = load_dataset("vbdai/TAD", name="exact_answer_action") mc_action =… See the full description on the dataset page: https://huggingface.co/datasets/vbdai/TAD.