pamasan/duckad-data
DuckAD Driving Dataset Expert driving demonstrations for DuckAD, an end-to-end vision-based driving model, collected in CARLA on custom Duckietown-style maps. A rule-aware expert driver was rolled out under six traffic/obstacle scenarios; every frame pairs a front camera image with the expert's future trajectory, a high-level navigation command, and ground-truth bird's-eye-view (BEV) semantics. 214,200 training frames from two maps (duckietown_04, duckietown_05), six scenarios… See the full description on the dataset page: https://huggingface.co/datasets/pamasan/duckad-data.
Update README.md
Update README.md
drop CARLA citation block
license section: custom assets, CARLA citation as recommendation not obligation
usage: show all modalities (seg, bev, command, speed, temporal trajectory)
dataset card: stats, schema, scenarios, license
Upload dataset
remove 20-row smoke-test splits before full upload
Upload dataset
initial commit
