Seeing double without confusion: Structure-from-motion in highly ambiguous scenes.
Nianjuan JiangPing TanLoong Fah CheongPublished in: CVPR (2012)
Keyphrases
- structure from motion
- highly ambiguous
- computer vision
- urban scenes
- bundle adjustment
- calibrated images
- word sense disambiguation
- natural language
- planar scenes
- camera parameters
- image sequences
- point cloud
- missing data
- dynamic scenes
- camera motion
- feature tracking
- scene structure
- reconstruction from multiple images
- focal length
- camera positions
- reprojection error
- image based localization
- real scenes
- multiple objects
- camera pose
- motion estimation
- information extraction
- video sequences