RUST: Latent Neural Scene Representations from Unposed Imagery.
Mehdi S. M. SajjadiAravindh MahendranThomas KipfEtienne PotDaniel DuckworthMario LucicKlaus GreffPublished in: CoRR (2022)
Keyphrases
- d scene
- distributed representations
- network architecture
- neural network
- scene analysis
- object models
- single image
- three dimensional
- video sequences
- imaging sensors
- real scenes
- high resolution
- latent variables
- computer vision
- scene classification
- man made
- image sequences
- dynamic scenes
- visual motion
- depth map
- moving objects
- scene images
- outdoor scenes
- observed scene
- multiple objects
- scene understanding
- complex scenes
- multiple images
- spatial relations
- higher level
- input image
- neural model
- scene structure
- image regions
- image processing
- street view