当前位置: X-MOL 学术Robot. Intell. Autom. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
MDRNet: a lightweight network for real-time semantic segmentation in street scenes
Robotic Intelligence and Automation ( IF 1.9 ) Pub Date : 2021-10-25 , DOI: 10.1108/aa-06-2021-0078
Yingpeng Dai 1 , Junzheng Wang 1 , Jiehao Li 1 , Jing Li 1
Affiliation  

Purpose

This paper aims to focus on the environmental perception of unmanned platform under complex street scenes. Unmanned platform has a strict requirement both on accuracy and inference speed. So how to make a trade-off between accuracy and inference speed during the extraction of environmental information becomes a challenge.

Design/methodology/approach

In this paper, a novel multi-scale depth-wise residual (MDR) module is proposed. This module makes full use of depth-wise separable convolution, dilated convolution and 1-dimensional (1-D) convolution, which is able to extract local information and contextual information jointly while keeping this module small-scale and shallow. Then, based on MDR module, a novel network named multi-scale depth-wise residual network (MDRNet) is designed for fast semantic segmentation. This network could extract multi-scale information and maintain feature maps with high spatial resolution to mitigate the existence of objects at multiple scales.

Findings

Experiments on Camvid data set and Cityscapes data set reveal that the proposed MDRNet produces competitive results both in terms of computational time and accuracy during inference. Specially, the authors got 67.47 and 68.7% Mean Intersection over Union (MIoU) on Camvid data set and Cityscapes data set, respectively, with only 0.84 million parameters and quicker speed on a single GTX 1070Ti card.

Originality/value

This research can provide the theoretical and engineering basis for environmental perception on the unmanned platform. In addition, it provides environmental information to support the subsequent works.



中文翻译:

MDRNet:用于街景实时语义分割的轻量级网络

目的

本文旨在关注复杂街景下无人平台的环境感知。无人平台对准确性和推理速度都有严格的要求。那么如何在环境信息的提取过程中,在准确率和推理速度之间进行权衡就成为一个挑战。

设计/方法/方法

在本文中,提出了一种新颖的多尺度深度残差(MDR)模块。该模块充分利用了深度可分离卷积、扩张卷积和一维(1-D)卷积,在保持该模块小规模和浅层的同时,能够联合提取局部信息和上下文信息。然后,基于 MDR 模块,设计了一种名为多尺度深度残差网络 (MDRNet) 的新型网络,用于快速语义分割。该网络可以提取多尺度信息并维护具有高空间分辨率的特征图,以减轻多尺度对象的存在。

发现

在 Camvid 数据集和 Cityscapes 数据集上的实验表明,所提出的 MDRNet 在推理过程中的计算时间和准确性方面都产生了有竞争力的结果。特别是,作者在 Camvid 数据集和 Cityscapes 数据集上分别获得了 67.47% 和 68.7% 的 Mean Intersection over Union (MIoU),在单张 GTX 1070Ti 卡上只有 84 万个参数和更快的速度。

原创性/价值

该研究可为无人平台环境感知提供理论和工程依据。此外,它还提供环境信息以支持后续工作。

更新日期:2021-11-23
down
wechat
bug