Decision fusion for urine particle classification in multispectral images
Proceedings of the Seventh Indian Conference on Computer Vision, Graphics and Image Processing
Action recognition with appearance-motion features and fast search trees
Computer Vision and Image Understanding
Non-sparse multiple kernel fisher discriminant analysis
The Journal of Machine Learning Research
Self-taught dimensionality reduction on the high-dimensional small-sized data
Pattern Recognition
Efficient discriminative projections for compact binary descriptors
ECCV'12 Proceedings of the 12th European conference on Computer Vision - Volume Part I
Per-patch descriptor selection using surface and scene properties
ECCV'12 Proceedings of the 12th European conference on Computer Vision - Volume Part VI
A picture is worth a thousand tags: automatic web based image tag expansion
ACCV'12 Proceedings of the 11th Asian conference on Computer Vision - Volume Part II
Global plus local: A complete framework for feature extraction and recognition
Pattern Recognition
Hi-index | 0.14 |
In this paper, we present Linear Discriminant Projections (LDP) for reducing dimensionality and improving discriminability of local image descriptors. We place LDP into the context of state-of-the-art discriminant projections and analyze its properties. LDP requires a large set of training data with point-to-point correspondence ground truth. We demonstrate that training data produced by a simulation of image transformations leads to nearly the same results as the real data with correspondence ground truth. This makes it possible to apply LDP as well as other discriminant projection approaches to the problems where the correspondence ground truth is not available, such as image categorization. We perform an extensive experimental evaluation on standard data sets in the context of image matching and categorization. We demonstrate that LDP enables significant dimensionality reduction of local descriptors and performance increases in different applications. The results improve upon the state-of-the-art recognition performance with simultaneous dimensionality reduction from 128 to 30.