This paper presents a fast algorithm for restoring video sequences. The proposed algorithm, as opposed to existing methods, does not consider video restoration as a sequence of image restoration problems. Rather, it treats a video sequence as a space-time volume and poses a space-time total variation regularization to enhance the smoothness of the solution. The optimization problem is solved by transforming the original unconstrained minimization problem to an equivalent constrained minimization problem. An augmented Lagrangian method is used to handle the constraints, and an alternating direction method is used to iteratively find solutions to the subproblems. The proposed algorithm has a wide range of applications, including video deblurring and denoising, video disparity refinement, and hot-air turbulence effect reduction.
Download full-text PDF |
Source |
|---|---|
| http://dx.doi.org/10.1109/TIP.2011.2158229 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
A finite volume adaptation of beam-to-beam contact interactions implemented for geometrically exact Simo-Reissner beams.
Comput Mech
June 2024
School of Mechanical and Materials Engineering, University College Dublin, Dublin, Ireland.
This paper presents an adaption of the finite-element based beam-to-beam contact interactions into a finite volume numerical framework. A previous work of the same authors, where a cell-centred based finite volume implementation of geometrically exact nonlinear Simo-Reissner beams was developed, is used as an underlying mathematical model. An implicit contact procedure is developed for both point-to-point and line-to-line beam frictionless contact interactions, and is implemented using the cell-centred finite volume method.
View Article and Find Full Text PDFAn approach for load frequency control enhancement in two-area hydro-wind power systems using LSTM + GA-PID controller with augmented lagrangian methods.
Sci Rep
January 2025
Department of Theoretical Electrical Engineering and Diagnostics of Electrical Equipment, Institute of Electrodynamics, National Academy of Sciences of Ukraine, Beresteyskiy, 56, Kyiv-57, Kyiv, 03680, Ukraine.
This paper proposes an advanced Load Frequency Control (LFC) strategy for two-area hydro-wind power systems, using a hybrid Long Short-Term Memory (LSTM) neural network combined with a Genetic Algorithm-optimized PID (GA-PID) controller. Traditional PID controllers, while extensively used, often face limitations in handling the nonlinearities and uncertainties inherent in interconnected power systems, leading to slower settling time and higher overshoot during load disturbances. The LSTM + GA-PID controller mitigates these issues by utilizing LSTM's capacity to learn from historical data by using gradient descent to forecast the future disturbances, while the GA optimizes the PID parameters in real time, ensuring dynamic adaptability and improved control precision.
View Article and Find Full Text PDFDirect Givens rotation method based on error back-propagation algorithm for self-consistent field solution.
J Chem Phys
January 2025
Department of Chemistry and Biochemistry, School of Advanced Science and Engineering, Waseda University, 3-4-1 Okubo, Shinjuku, Tokyo 169-8555, Japan.
The self-consistent field (SCF) procedure is the standard technique for solving the Hartree-Fock and Kohn-Sham density functional theory calculations, while convergence is not theoretically guaranteed. Direct minimization methods, such as the augmented Lagrangian method (ALM) and second-order SCF (SOSCF), obtain the SCF solution by minimizing the Lagrangian with the gradient. In SOSCF, molecular orbitals are optimized by truncating the Taylor expansion of a unitary matrix represented in exponential form to ensure the orthonormality condition.
View Article and Find Full Text PDFDistributed nonconvex optimization subject to globally coupled constraints via collaborative neurodynamic optimization.
Neural Netw
December 2024
School of Mathematics and Statistics, Yili Normal University, Yining 835000, China.
In this paper, a recurrent neural network is proposed for distributed nonconvex optimization subject to globally coupled (in)equality constraints and local bound constraints. Two distributed optimization models, including a resource allocation problem and a consensus-constrained optimization problem, are established, where the objective functions are not necessarily convex, or the constraints do not guarantee a convex feasible set. To handle the nonconvexity, an augmented Lagrangian function is designed, based on which a recurrent neural network is developed for solving the optimization models in a distributed manner, and the convergence to a local optimal solution is proven.
View Article and Find Full Text PDFScalable log-ratio lasso regression for enhanced microbial feature selection with FLORAL.
Cell Rep Methods
November 2024
Department of Hematology and Hematopoietic Cell Transplantation, City of Hope National Medical Center, Los Angeles, CA, USA; Hematologic Malignancies Research Institute, City of Hope National Medical Center, Los Angeles, CA, USA; Comprehensive Cancer Center, City of Hope National Medical Center, Los Angeles, CA, USA. Electronic address:
Identifying predictive biomarkers of patient outcomes from high-throughput microbiome data is of high interest, while existing computational methods do not satisfactorily account for complex survival endpoints, longitudinal samples, and taxa-specific sequencing biases. We present FLORAL, an open-source tool to perform scalable log-ratio lasso regression and microbial feature selection for continuous, binary, time-to-event, and competing risk outcomes, with compatibility for longitudinal microbiome data as time-dependent covariates. The proposed method adapts the augmented Lagrangian algorithm for a zero-sum constraint optimization problem while enabling a two-stage screening process for enhanced false-positive control.
View Article and Find Full Text PDFWant AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!