Down to Earth: Using Semantics for Robust Hypothesis Selection for the Five-Point Algorithm.
Andreas KuhnTrue PriceJan-Michael FrahmHelmut MayerPublished in: GCPR (2017)
Keyphrases
- theoretical analysis
- detection algorithm
- computationally efficient
- improved algorithm
- computational complexity
- parameter tuning
- selection algorithm
- learning algorithm
- recognition algorithm
- dynamic programming
- computational cost
- high accuracy
- simulated annealing
- matching algorithm
- image processing
- similarity measure
- objective function
- experimental evaluation
- np hard
- classification algorithm
- highly efficient
- decision trees
- times faster
- kalman filter
- significant improvement
- worst case
- cost function
- graph cuts
- particle swarm optimization
- linear programming
- probabilistic model
- markov random field