Sinkhorn Flow as Mirror Flow: A Continuous-Time Framework for Generalizing the Sinkhorn Algorithm.
Mohammad Reza KarimiYa-Ping HsiehAndreas KrausePublished in: AISTATS (2024)
Keyphrases
- detection algorithm
- optimal solution
- preprocessing
- probabilistic model
- significant improvement
- computational cost
- experimental evaluation
- dynamic programming
- optimization algorithm
- high accuracy
- recognition algorithm
- np hard
- cost function
- objective function
- learning algorithm
- flow patterns
- maximum flow
- neural network
- convergence rate
- bayesian framework
- times faster
- matching algorithm
- min cost
- theoretical analysis
- particle swarm optimization
- linear programming
- state space
- segmentation algorithm
- computationally efficient
- expectation maximization
- clustering method
- input data
- worst case
- field of view
- denoising
- multiple instance learning
- graph based algorithm
- k means
- computational complexity