Sciweavers

182 search results - page 9 / 37
» Neural Network Ensembles from Training Set Expansions
Sort
View
IJCNN
2006
IEEE
14 years 1 months ago
Oscillatory Network for Synchronization-Based Adaptive Image Segmentation
— Oscillatory network model with controllable oscillator dynamics and self-organized dynamical coupling has been created for synchronization-based image processing. The model was...
Eugene Grichuk, Margarita Kuzmina, Edward A. Manyk...
ICDAR
2003
IEEE
14 years 1 months ago
Best Practices for Convolutional Neural Networks Applied to Visual Document Analysis
Neural networks are a powerful technology for classification of visual inputs arising from documents. However, there is a confusing plethora of different neural network methods th...
Patrice Simard, David Steinkraus, John C. Platt
ECCV
2008
Springer
14 years 9 months ago
Training Hierarchical Feed-Forward Visual Recognition Models Using Transfer Learning from Pseudo-Tasks
Abstract. Building visual recognition models that adapt across different domains is a challenging task for computer vision. While feature-learning machines in the form of hierarchi...
Amr Ahmed, Kai Yu, Wei Xu, Yihong Gong, Eric P. Xi...
MLDM
2001
Springer
14 years 9 days ago
Local Learning Framework for Recognition of Lowercase Handwritten Characters
Abstract. This paper proposes a general local learning framework to effectively alleviate the complexities of classifier design by means of “divide and conquer” principle and ...
Jian-xiong Dong, Adam Krzyzak, Ching Y. Suen
ANNPR
2006
Springer
13 years 11 months ago
Fast Training of Linear Programming Support Vector Machines Using Decomposition Techniques
Abstract. Decomposition techniques are used to speed up training support vector machines but for linear programming support vector machines (LP-SVMs) direct implementation of decom...
Yusuke Torii, Shigeo Abe