Research papers · Posted 2026.05.13 · Publication / acceptance: author-reported
Generative Texture Diversification of 3D Pedestrians for Robust Autonomous Driving Perception
Generative Texture Diversification of 3D Pedestrians for Robust Autonomous Driving Perception
Overview
The study diversifies pedestrian textures using StyleGAN2 and maps them onto 3D assets for synthetic training data. Experiments examine 2D detection robustness and the geometric domain gaps affecting 3D perception.
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.
Published at SAIAD 2026 Workshop at CVPR 2026
Version: arXiv v1 — manuscript revision