Research Article
Lightweight Neural Network for Sketch Recognition on Mobile Phones
@INPROCEEDINGS{10.1007/978-3-030-51005-3_35, author={Ni Kong and Hong Hou and Zhe Bai and Xiaoqun Guo}, title={Lightweight Neural Network for Sketch Recognition on Mobile Phones}, proceedings={Science and Technologies for Smart Cities. 5th EAI International Summit, SmartCity360, Braga, Portugal, December 4-6, 2019, Proceedings}, proceedings_a={SMARTCITY}, year={2020}, month={7}, keywords={Sketch recognition Lightweight networks Depthwise separable convolution}, doi={10.1007/978-3-030-51005-3_35} }- Ni Kong
Hong Hou
Zhe Bai
Xiaoqun Guo
Year: 2020
Lightweight Neural Network for Sketch Recognition on Mobile Phones
SMARTCITY
Springer
DOI: 10.1007/978-3-030-51005-3_35
Abstract
With the popularity of smart terminals, people tend to draw simple sketches to express emotions and ideas in communication, which means the era of reading pictures is coming. Therefore, in the field of sketch recognition, the application of deep network models in mobile devices is an irreversible trend. However, most existing works with good performance has a large number of parameters by using deep learning method. In order to further improve recognition speed and ensure the accuracy, we propose a lightweight neural network architecture to recognize sketch object. Specifically, we apply depthwise separable convolution into the network to reduce parameters and adjust the network effectively for the sparsity of sketch. Outperforming the state-of-the-art approaches, we achieves 85.3% and 83.7% on TU-Berlin and QuickDraw benchmarks respectively. Furthermore, the number of parameters is reduced to a large extent, which are 5% and 20% of the amount of Sketch-A-Net and MobileNets. We also develop a sketch recognition application for mobile phones to demonstrate the proposed scheme.
