Dual Space Preconditioning for Gradient Descent.
Chris J. MaddisonDaniel PaulinYee Whye TehArnaud DoucetPublished in: SIAM J. Optim. (2021)
Keyphrases
- dual space
- conjugate gradient
- iterative methods
- cost function
- training algorithm
- edge preserving
- loss function
- maximum likelihood estimation
- convergence rate
- objective function
- early vision
- optical flow computation
- primal dual
- faster convergence
- learning algorithm
- levenberg marquardt
- inverse problems
- pairwise
- image sequences