Search

Abstract Face recognition is still a hot topic under investigation due to many challenges of variation including the difference in poses, illumination, expression, occlusion and scenes. Recently, deep learning methods achieved remarkable results in image representation and recognition fields. Such methods extract salient features automatically from images to reduce the dimension and obtain more useful representation of raw data. In this paper, the proposed face recognition system namely Deep Stacked Denoising Sparse Autoencoders (DSDSA) combines the deep neural network technology, sparse autoencoders and denoising task. Autoencoder is used to construct a neural network that learns an approximation of an identity function by placing constraints to learn fine representations of the inputs. Autoencoders have unique capabilities in dealing with interpretation of the input data; in this way produce more meaningful results. They are successfully applied to many object recognition fields. For the classification task, two classifiers were used, namely multi-class SVM and Softmax classifier. Experimental results on known face databases including ORL, Yale, Caltech and a subset of PubFig show that the proposed system yields promising performance and achieves comparable accuracy. [ABSTRACT FROM AUTHOR]

This paper deals with site selection problems for wind power plants and aims to propose a structural procedure for determining the most feasible sites. The application area is Western Turkey. The methodology is mainly composed of two stages: the first stage is pre-elimination of infeasible sites, and the second stage is evaluation of the available ones. Geographic Information Systems (GIS) are used to generate layers of data and to apply the elimination criteria and constraints. The alternative land areas are handled in terms of identical-sized grids, which are large enough to install one wind turbine each. Multiple Criteria Decision Analysis (MCDA) is then used to rank and sort the grids via the identified evaluation criteria. The problem is evaluated in 13 fields, which are a collection of several grids in order to evaluate larger scale areas to construct wind farms rather than for individual turbines. The evaluations are made both at grid (micro) and field (macro) levels and both deterministic and uncertain data are used. The results reveal a high level of consensus on the most feasible sites between the different MCDA methods applied. The proposed methodology provides a structured decision aid, which can be applied to other energy site selection problems. [ABSTRACT FROM AUTHOR]