Cifar-10 Image Classification with Convolutional Neural Networks for Embedded Systems

Convolutional Neural Networks (CNN) have been successfully applied to image classification problems. Although powerful, they require a large amount of memory. The purpose of this paper is to perform image classification using CNNs on the embedded systems, where only a limited amount of memory is available. Our experimental analysis shows that 85.9% image classification accuracy is obtained by our framework while requiring 2GB memory only, making our framework ideal to be used in embedded systems.

[1]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[2]  Xiaolin Hu,et al.  Recurrent convolutional neural network for object recognition , 2015, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[3]  M. Fatih Demirci,et al.  Age Classification Using an Optimized CNN Architecture , 2017, ICCDA '17.

[4]  M. Fatih Demirci,et al.  Comparison of Three Different CNN Architectures for Age Classification , 2017, 2017 IEEE 11th International Conference on Semantic Computing (ICSC).