CLIP-FO3D: Learning Free Open-world 3D Scene Representations from 2D Dense CLIP.

Junbo Zhang Runpei Dong Kaisheng Ma

Published in: ICCV (Workshops) (2023)

Keyphrases

d scene
learning process
active learning
scene flow
open world
machine learning
computer vision
image sequences
viewpoint
depth map
incomplete information
image based rendering