MDSNet: A lightweight network for real-time vision task on the unmanned mobile robot

Yingpeng Dai, Junzheng Wang, Jing Li*

*此作品的通讯作者

科研成果: 书/报告/会议事项章节会议稿件同行评审

摘要

To makes a trade-off between accuracy and inference speed for semantic segmentation, Multi-Scale Depthwise Separation network (MDSNet) is designed to be effective both in terms of accuracy and inference speed. This network extract local information and contextual information jointly and has feature maps with high spatial resolution. Compared with state-of-the-art algorithms, MDSNet achieves 66.57 MIoU on Camvid with only 0.5M parameters and 79.4 FPS inference speed on a single GTX 1070Ti card. Besides, MDS is deployed on the unmanned platform to test performance under different conditions. The results show that the proposed algorithm performs well on real-time applications in the real world.

源语言英语
主期刊名Proceedings - 2022 International Symposium on Intelligent Robotics and Systems, ISoIRS 2022
出版商Institute of Electrical and Electronics Engineers Inc.
21-25
页数5
ISBN(电子版)9781665455213
DOI
出版状态已出版 - 2022
活动2022 International Symposium on Intelligent Robotics and Systems, ISoIRS 2022 - Virtual, Online, 中国
期限: 14 10月 202216 10月 2022

出版系列

姓名Proceedings - 2022 International Symposium on Intelligent Robotics and Systems, ISoIRS 2022

会议

会议2022 International Symposium on Intelligent Robotics and Systems, ISoIRS 2022
国家/地区中国
Virtual, Online
时期14/10/2216/10/22

指纹

探究 'MDSNet: A lightweight network for real-time vision task on the unmanned mobile robot' 的科研主题。它们共同构成独一无二的指纹。

引用此