GPSlam: Marrying Sparse Geometric and Dense Probabilistic Visual Mapping.
Katrin PirkerMatthias RütherGerald SchweighoferHorst BischofPublished in: BMVC (2011)
Keyphrases
- dense stereo
- motion field estimation
- bayesian networks
- visual perception
- high dimensional
- probabilistic model
- dense optical flow
- visual information
- dense reconstruction
- probabilistic logic
- sparse data
- shape recovery
- dense motion estimation
- compressed sensing
- geometric information
- uncertain data
- visual features
- multi view
- neural network
- visual cues
- sparse representation
- dense sampling
- object recognition