Advanced machine learning algorithms for high-dimensional remote sensing image processing
Event details
| Date | 26.04.2010 |
| Hour | 16:15 |
| Speaker | Tuia, Devis (Université de Lausanne-UNIL) |
| Location |
GR B3 30
|
| Category | Conferences - Seminars |
The technical developments in recent years have brought the quantity and quality of digital information to an unprecedented level, as enormous archives of satellite images are available to the users. However, even if these advances open more and more possibilities in the use of digital imagery, they also rise several problems of storage and treatment. The latter is considered in this presentation: the processing of very high spatial and spectral resolution images is treated with approaches based on data-driven algorithms relying on kernel methods.
In particular, the problem of image classification, i.e. the categorization of the image's pixels into a reduced number of classes reflecting spectral and contextual properties, is studied through the different models presented. The accent is put on algorithmic efficiency and the simplicity of the approaches proposed, to avoid too complex models that would not be used by users.
After a short introduction on machine learning for remote sensing, two problems will be considered : first, the question of high dimensionality and collinearity of the image features is studied by an adaptive model learning the relevant image features. This model provides automatically an accurate classifier and a ranking of the relevance of the single features. Second is the question of scarcity and unreliability of the labeled information: when confronted to such situations, the user can either construct the labeled set iteratively by direct interaction with the machine (active learning) or use the unlabeled data to increase robustness and quality of the description of data (semi-supervised learning). Both solutions will be discussed during the presentation and advantages and limitations of the different strategies will be pointed out.
Links
Practical information
- General public
- Free