MBO method for image processing and classification using a graphical framework
Event details
| Date | 03.07.2015 |
| Hour | 14:00 |
| Speaker | Dr. Ekaterina Merkurjev, UCLA |
| Location |
INF119
|
| Category | Conferences - Seminars |
We present a graph-based algorithm for image processing and classification of high-dimensional data. The semi-supervised method uses a graph adaptation of the classical numerical Merriman-Bence-Osher (MBO) scheme, and can be extended to the multiclass case via the Gibbs simplex. We show examples of the application of the algorithm in the areas of image inpainting, image segmentation and object detection using hyperspectral video sequences.
Bio: Solid background in applied and computational mathematics, differential equations, numerical analysis, optimization, scientific computing.
Variational and PDE-based methods for machine learning, data analysis, and image processing using a graphical framework. Applications include classification of high-dimensional data, image segmentation and image inpainting.
Bio: Solid background in applied and computational mathematics, differential equations, numerical analysis, optimization, scientific computing.
Variational and PDE-based methods for machine learning, data analysis, and image processing using a graphical framework. Applications include classification of high-dimensional data, image segmentation and image inpainting.
Practical information
- General public
- Free
Organizer
- Signal Processing Laboratory (LTS2)
Contact
- Xavier Bresson