DreamScene4D: Dynamic Multi-Object Scene Generation from Monocular Videos.

Wen-Hsuan Chu Lei Ke Katerina Fragkiadaki

Published in: CoRR (2024)

Keyphrases

multi object
multiple objects
video sequences
image sequences
dynamic scenes
moving camera
human perception
video frames
motion trajectories
optical flow
multiple images
video data
spatio temporal
d scene
three dimensional
pose estimation
video analysis
single image
high quality
depth cues