In this paper, we explore the redundancy in convolutional neural network, which scales with the complexity of vision tasks. Considering that many front-end visual systems are interested in only a limited range of visual targets, the removing of task-specified network redundancy can promote a wide range of potential applications. We propose a task-specified knowledge distillation algorithm to derive a simplified model with pre-set computation cost and minimized accuracy loss, which suits the resource constraint front-end systems well. Experiments on the MNIST and CIFAR10 datasets demonstrate the feasibility of the proposed approach as well as the existence of task-specified redundancy.
@article{arxiv.1609.05695,
title = {A scalable convolutional neural network for task-specified scenarios via knowledge distillation},
author = {Mengnan Shi and Fei Qin and Qixiang Ye and Zhenjun Han and Jianbin Jiao},
journal= {arXiv preprint arXiv:1609.05695},
year = {2017}
}