TITLE

Unsupervised Learning of Overlapping Image Components Using Divisive Input Modulation

AUTHOR(S)
Spratling, M. W.; DeMeyer, K.; Kompass, R.
PUB. DATE
January 2009
SOURCE
Computational Intelligence & Neuroscience;2009, Special section p1
SOURCE TYPE
Academic Journal
DOC. TYPE
Article
ABSTRACT
This paper demonstrates that nonnegative matrix factorisation is mathematically related to a class of neural networks that employ negative feedback as a mechanism of competition. This observation inspires a novel learning algorithm which we call Divisive Input Modulation (DIM). The proposed algorithm provides a mathematically simple and computationally efficient method for the unsupervised learning of image components, even in conditions where these elementary features overlap considerably. To test the proposed algorithm, a novel artificial task is introduced which is similar to the frequently-used bars problem but employs squares rather than bars to increase the degree of overlap between components. Using this task, we investigate how the proposed method performs on the parsing of artificial images composed of overlapping features, given the correct representation of the individual components; and secondly, we investigate how well it can learn the elementary components from artificial training images. We compare the performance of the proposed algorithm with its predecessors including variations on these algorithms that have produced state-of-the-art performance on the bars problem. The proposed algorithm is more successful than its predecessors in dealing with overlap and occlusion in the artificial task that has been used to assess performance.
ACCESSION #
47715136

 

Related Articles

  • A fuzzy neighborhood-based training algorithm for feedforward neural networks. Nasr, Mounir; Chtourou, Mohamed // Neural Computing & Applications;2009, Vol. 18 Issue 2, p127 

    In this work we present a new hybrid algorithm for feedforward neural networks, which combines unsupervised and supervised learning. In this approach, we use a Kohonen algorithm with a fuzzy neighborhood for training the weights of the hidden layers and gradient descent method for training the...

  • EVOLVING CONNECTION WEIGHTS FOR PATTERN STORAGE AND RECALL IN HOPFIELD MODEL OF FEEDBACK NEURAL NETWORKS USING A GENETIC ALGORITHM. Singh, T. P.; Jabin, Suraiya // International Journal on Soft Computing;May2012, Vol. 3 Issue 2, p55 

    In this paper, implementation of a genetic algorithm has been described to store and later, recall of some prototype patterns in Hopfield neural network associative memory. Various operators of genetic algorithm (mutation, cross-over, elitism etc) are used to evolve the population of optimal...

  • Soft Computing Using Neural Estimation with LMI-Based Model Transformation for OMR-Based Control of the Buck Converter. Al-Rabadi, Anas N.; Alsmadi, Othman M. K. // Engineering Letters;2009, Vol. 17 Issue 2, p101 

    This paper introduces a new method of intelligent control to control the Buck converter using newly developed small signal model of the pulse width modulation (PWM) switch. The new method uses recurrent supervised neural network to estimate certain parameters of the transformed system matrix...

  • An Intelligent Offline Handwriting Recognition System Using Evolutionary Neural Learning Algorithm and Rule Based Over Segmented Data Points. Ghosh, Ranadhir; Ghosh, Moumita // Journal of Research & Practice in Information Technology;Feb2005, Vol. 37 Issue 1, p73 

    In this paper we propose a novel technique of using a hybrid evolutionary method, which uses combination of genetic algorithm and matrix based solution methods such as QR factorization. The training of the model is based on a layer based hierarchical structure for the architecture and the...

  • A New Optimized GA-RBF Neural Network Algorithm. Weikuan Jia; Dean Zhao; Tian Shen; Chunyang Su; Chanli Hu; Yuyan Zhao // Computational Intelligence & Neuroscience;2014, Vol. 2015, p1 

    When confronting the complex problems, radial basis function (RBF) neural network has the advantages of adaptive and selflearning ability, but it is difficult to determine the number of hidden layer neurons, and the weights learning ability from hidden layer to the output layer is low; these...

  • A Study of Driver's Route Choice Behavior Based on Evolutionary Game Theory. Xiaowei Jiang; Yanjie Ji; Muqing Du; Wei Deng // Computational Intelligence & Neuroscience;2014, Vol. 2015, p1 

    This paper proposes a route choice analytic method that embeds cumulative prospect theory in evolutionary game theory to analyze how the drivers adjust their route choice behaviors under the influence of the traffic information. A simulated network with two alternative routes and one variable...

  • Transfer Learning via Inter-Task Mappings for Temporal Difference Learning. Taylor, Matthew E.; Stone, Peter; Yaxin Liu // Journal of Machine Learning Research;9/1/2007, Vol. 8 Issue 9, p2125 

    Temporal difference (TD) learning (Sutton and Barto, 1998) has become a popular reinforcement learning technique in recent years. TD methods, relying on function approximators to generalize learning to novel situations, have had some experimental successes and have been shown to exhibit some...

  • Comparison of Diverse Ensemble Neural Network for Large Data Classification. Mohamad, Mumtazimah; Saman, Md Yazid Mohd // International Journal of Advances in Soft Computing & Its Applic;Nov2015, Vol. 7 Issue 3, p67 

    In a large dataset classification, a higher number of attributes commonly evolve over time, where many dynamic learning strategies have been proposed such as the ensemble network and incremental neural network. Ensemble network is a learning paradigm where many neural networks are jointly used...

  • Particle swarm optimization for ensembling generation for evidential k-nearest-neighbour classifier. Nanni, Loris; Lumini, Alessandra // Neural Computing & Applications;2009, Vol. 18 Issue 2, p105 

    The problem addressed in this paper concerns the ensembling generation for evidential k-nearest-neighbour classifier. An efficient method based on particle swarm optimization (PSO) is here proposed. We improve the performance of the evidential k-nearest-neighbour (EkNN) classifier using a random...

Share

Read the Article

Courtesy of VIRGINIA BEACH PUBLIC LIBRARY AND SYSTEM

Sorry, but this item is not currently available from your library.

Try another library?
Sign out of this library

Other Topics