Efficiently locating photographs in many panoramas
Proceedings of the 18th SIGSPATIAL International Conference on Advances in Geographic Information Systems
Face recognition based on the multi-scale local image structures
Pattern Recognition
Towards reliable matching of images containing repetitive patterns
Pattern Recognition Letters
International Journal of Computer Vision
Face recognition using the POEM descriptor
Pattern Recognition
Thick boundaries in binary space and their influence on nearest-neighbor search
Pattern Recognition Letters
Efficient discriminative projections for compact binary descriptors
ECCV'12 Proceedings of the 12th European conference on Computer Vision - Volume Part I
Descriptor learning using convex optimisation
ECCV'12 Proceedings of the 12th European conference on Computer Vision - Volume Part I
Detecting and reconstructing 3d mirror symmetric objects
ECCV'12 Proceedings of the 12th European conference on Computer Vision - Volume Part II
Per-patch descriptor selection using surface and scene properties
ECCV'12 Proceedings of the 12th European conference on Computer Vision - Volume Part VI
A feature compression scheme for large scale image retrieval systems
Proceedings of the 27th Conference on Image and Vision Computing New Zealand
Image and Vision Computing
Proceedings of the 3rd ACM conference on International conference on multimedia retrieval
An improvement to the SIFT descriptor for image representation and matching
Pattern Recognition Letters
Large-scale Structure-from-Motion Reconstruction with small memory consumption
Proceedings of International Conference on Advances in Mobile Computing & Multimedia
IJCAI'13 Proceedings of the Twenty-Third international joint conference on Artificial Intelligence
Hi-index | 0.14 |
In this paper, we explore methods for learning local image descriptors from training data. We describe a set of building blocks for constructing descriptors which can be combined together and jointly optimized so as to minimize the error of a nearest-neighbor classifier. We consider both linear and nonlinear transforms with dimensionality reduction, and make use of discriminant learning techniques such as Linear Discriminant Analysis (LDA) and Powell minimization to solve for the parameters. Using these techniques, we obtain descriptors that exceed state-of-the-art performance with low dimensionality. In addition to new experiments and recommendations for descriptor learning, we are also making available a new and realistic ground truth data set based on multiview stereo data.