A Gradient-Based Algorithm to Deceive Deep Neural Networks.
Tianying XieYantao LiPublished in: ICONIP (4) (2019)
Keyphrases
- neural network
- learning algorithm
- worst case
- high accuracy
- computationally efficient
- linear programming
- dynamic programming
- objective function
- optimal solution
- k means
- improved algorithm
- times faster
- optimization algorithm
- computational complexity
- computational cost
- preprocessing
- search space
- np hard
- particle swarm optimization
- theoretical analysis
- cost function
- path planning
- significant improvement
- face recognition
- segmentation algorithm
- detection algorithm
- clustering method
- convex hull
- convergence rate
- recognition algorithm