arrow
Return

MSTDNet: Multi-scale traffic object detection network with smooth information perception

delete2025-10-12
delete0
PRE
AI
J
Jie Hua
王中元 (Zhongyuan Wang)
常军 cover
常军 (Jun Chang)
H
Hua Zou
马佳义 cover
马佳义 (Jiayi Ma)
DOI:10.1016/j.patcog.2025.112579delete
deleteOriginal
deleteOriginal request for help
deleteShare
deleteSave
Abstract

Abstract

En 中文
• Convolutions with larger controllable receptive fields are proposed. • A smooth multi-scale information perception module is proposed. • A lightweight interaction branch for improving performance is proposed. • MSTDNet performs optimally in marine, land, and foggy traffic scenes.

Journal

Pattern Recognition cover
Pattern Recognition
IF:
7.6
Papers:
1.3W
Citations:
4.5W

Organization

T
Tarim University
Scholars:
3.6K
Papers: 1.6K
Citations: 1.7K
W
wuhan university
Scholars:
8.0W
Papers: 5.8W
Citations: 70