GPU implementation of neural networks
School of Media, College of Information Science, Soongsil University, 1, SangDo-Dong, DongJak-Gu, Seoul, 156-743, Republic of Korea
Pattern Recognition, Vol. 37, No. 6. (June 2004), pp. 1311-1314
@article{oh2004gpu,
title={GPU implementation of neural networks},
author={Oh, K.S. and Jung, K.},
journal={Pattern Recognition},
volume={37},
number={6},
pages={1311–1314},
year={2004},
publisher={Elsevier}
}
Graphics processing unit (GPU) is used for a faster artificial neural network. It is used to implement the matrix multiplication of a neural network to enhance the time performance of a text detection system. Preliminary results produced a 20-fold performance enhancement using an ATI RADEON 9700 PRO board. The parallelism of a GPU is fully utilized by accumulating a lot of input feature vectors and weight vectors, then converting the many inner-product operations into one matrix operation. Further research areas include benchmarking the performance with various hardware and GPU-aware learning algorithms.
October 27, 2010 by hgpu
