Research papers · Posted 2022.06.01 · Publication / acceptance: author-reported
On the Choice of Data for Efficient Training and Validation of End-to-End Driving Models
On the Choice of Data for Efficient Training and Validation of End-to-End Driving Models
Overview
CARLA experiments examine training-data volume, generation choices, validation routes, and random seeds for end-to-end driving. The analysis investigates whether validation performance generalizes to unseen environments and whether reported improvements are statistically meaningful.
Publication and version information
Publication, acceptance or venue information is author-reported on arXiv. The exact source wording is preserved below.
Author-supplied source information
Author-supplied information on arXiv.
Accepted at CVPR VDU Workshop 2022
Version: arXiv v1 — manuscript revision