DyST: Towards Dynamic Neural Scene Representations on Real-World Videos.

Maximilian Seitzer Sjoerd van Steenkiste Thomas Kipf Klaus Greff Mehdi S. M. Sajjadi

Published in: ICLR (2024)

Keyphrases

real world
video sequences
dynamic textures
dynamic scenes
moving camera
data sets
d scene
case study

scene analysis
image sequences
single image
neighboring frames
video data
video images
network architecture

dynamic environments
moving objects
three dimensional
neural network
long video sequences
video sequences captured
text localization and recognition

distributed representations
video scene
visual patterns
video clips
video surveillance
human activities
input image