VisionPAD uses 3D Gaussian Splatting, self-supervised voxel velocity estimation, and photometric consistency to pre-train vision-centric driving models from images only.
pixelsplat: 3d gaussian splats from image pairs for scalable generalizable 3d reconstruction
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
VisionPAD: A Vision-Centric Pre-training Paradigm for Autonomous Driving
VisionPAD uses 3D Gaussian Splatting, self-supervised voxel velocity estimation, and photometric consistency to pre-train vision-centric driving models from images only.