datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
CVPR2026seqdataset-public
Auto-Annotation with Expert-Crafted Guidelines: A Study through 3D LiDAR Detection Benchmark
Inspired by the critical bottleneck of data annotation in autonomous driving and recent advancements in foundation models, this dataset introduces a novel evaluation paradigm: Auto-Annotation from Expert-Crafted Guidelines.
Unlike traditional 3D perception benchmarks that rely on massive amounts of annotated 3D point clouds for supervised learning, AutoExpert is designed to evaluate a… See the full description on the dataset page: https://huggingface.co/datasets/autoexpert-cvpr2026-workshop/dataset-public.