Abstract | ||
---|---|---|
The goal of this work is the design and implementation of a low-cost system-on-FPGA for handwritten digit recognition, based on a relatively deep and wide network of perceptrons. In order to increase the performance of the application on embedded processors whose performances are way below standard general purpose CPUs, a regularization method was used during the training phase of the neural network that allows for the drastic reduction of floating point operations. Our implementation achieves a 3 × speed-up toward a raw implementation without optimization, while keeping the accuracy in acceptable ranges. Our efforts reinforce the fact that FPGAs are suited for deploying complex artificial intelligence modules. |
Year | DOI | Venue |
---|---|---|
2015 | 10.1016/j.sysarc.2015.07.015 | Journal of Systems Architecture - Embedded Systems Design |
Keywords | Field | DocType |
Neural network,Soft processor,Regularization,Image processing | General purpose,Computer science,Floating point,Parallel computing,Field-programmable gate array,Image processing,Regularization (mathematics),Digit recognition,Artificial neural network,Perceptron,Embedded system | Journal |
Volume | Issue | ISSN |
61 | 10 | 1383-7621 |
Citations | PageRank | References |
0 | 0.34 | 6 |
Authors | ||
2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Luca Bochi Saldanha | 1 | 7 | 1.79 |
Christophe Bobda | 2 | 627 | 90.57 |