A neural network implementation of a saliency map model
β Scribed by Matthew de Brecht; Jun Saiki
- Publisher
- Elsevier Science
- Year
- 2006
- Tongue
- English
- Weight
- 788 KB
- Volume
- 19
- Category
- Article
- ISSN
- 0893-6080
No coin nor oath required. For personal study only.
π SIMILAR VOLUMES
The focus of this study is how we can efficiently implement the neural network backpropagation algorithm on a network of computers (NOC) for concurrent execution. We assume a distributed system with heterogeneous computers and that the neural network is replicated on each computer. We propose an arc
This paper presents a mapping scheme for the proposed implementation of neural network models on systolic arrays. The mapping technique is illustrated on the multilayer perceptron with back-propagation learning. Dependency graphs have been given that represent the operations in the execution phases