Towards Faster Training of Global Covariance Pooling Networks by Iterative Matrix Square Root Normalization.
Peihua LiJiangtao XieQilong WangZilin GaoPublished in: CoRR (2017)
Keyphrases
- square root
- floating point
- kalman filtering
- euclidean space
- arrival rate
- probability density function
- multi class
- covariance matrix
- visual tracking
- riemannian manifolds
- shape analysis
- metric space
- image sequences
- kernel methods
- fixed point
- sample size
- model selection
- support vector machine
- probability distribution
- video sequences