Parallelization of Direct-Forcing Immersed Boundary Method Using OpenACC.
Fang-An KuoShuen-Tai WangChau-Yi ChouYu-Bin FangPublished in: CANDAR Workshops (2019)
Keyphrases
- high accuracy
- synthetic data
- mutual information
- neural network
- detection method
- experimental evaluation
- feature vectors
- computational cost
- high precision
- classification method
- significant improvement
- theoretical analysis
- clustering method
- preprocessing
- computationally efficient
- fully automatic
- energy function
- data sets
- active contours
- training set
- pairwise
- classification accuracy
- cost function