Efficient data encoding for convolutional neural network application

Abstract : This article presents an approximate data encoding scheme called Significant Position Encoding (SPE). The encoding allows efficient implementation of the recall phase (forward propagation pass) of Convolutional Neural Networks (CNN)-a typical Feed-Forward Neural Network.This implementation uses only7bits data representation and achieves almost the same classification performance compared with the initial network: on MNIST handwriting recognition task, using this data encoding scheme losses only 0.03% in terms of recognition rate (99.27% vs. 99.3%). In terms of storage, we achieve a 12.5% gain compared with an 8 bits fixed-point implementation of the same CNN. Moreover, this data encoding allows efficient implementation of processing unit thanks to the simplicity of scalar product operation-the principal operation in a Feed-Forward Neural Network.
Document type :
Journal articles
Complete list of metadatas

https://hal-cea.archives-ouvertes.fr/cea-01846867
Contributor : Léna Le Roy <>
Submitted on : Monday, July 23, 2018 - 8:01:25 AM
Last modification on : Thursday, March 14, 2019 - 9:46:19 AM

Links full text

Identifiers

Collections

Citation

H.-P. Trinh, M. Duranton, M. Paindavoine. Efficient data encoding for convolutional neural network application. ACM Transactions on Architecture and Code Optimization, Association for Computing Machinery, 2014, 11 (4), ⟨10.1145/2685394⟩. ⟨cea-01846867⟩

Share

Metrics

Record views

74