arrow
返回

LiteCortexNet: toward efficient object detection at night

delete2022-07-07
delete4
PRE
AI
S
Sikai Wang
杨进 封面图
杨进 (Jin Yang)
D
Deng Chen *
黄晋 封面图
黄晋 (Jin Huang) *
Y
Yanduo Zhang
刘玮 封面图
刘玮 (Wei Liu)
Z
Zhaohui Zheng
Y
Yanan Li
DOI:10.1007/s00371-022-02560-9delete
delete原文链接
delete原文求助
delete分享
delete收藏
摘要

摘要

En 中文
Efficiently detecting objects in the complex background at night with low illumination remains a challenge for image processing. To address this issue, this paper proposes LiteCortexNet, a lightweight deep learning object detection model inspired by the visual cortex. The model performs intrinsic image decomposition end-to-end to obtain the illumination-independent reflection component, fuses it with the output result of the depth-wise separable convolutional encoder, and then, sends it to the lightweight detection head for object classification and positioning. Leveraging the channel-wise attention mechanism, our model is optimized for detecting small objects as well as obscured objects. In order to evaluate our method, an image dataset of railway maintenance tools was constructed. Experimental results show that the proposed model achieves 90.56% mAP at 66FPS on this dataset, which outperforms state-of-the-art object detection models such as YoloV4 (Bochkovskiy et al. in arXiv:2004.10934) (82.34% mAP at 45FPS).
Keyword:
Object detection
Retinex
Deep learning
Nighttime image

期刊

Visual Computer 封面图
Visual Computer
IF:
2.9
论文数:
4.6K
被引数:
6.5K

机构

W
wuhan textile university
学者数:
6.7K
论文数: 4.0K
被引数: 3
W
wuhan institute of technology
学者数:
1.0W
论文数: 6.6K
被引数: 11
引用论文

引用论文

err分享
err收藏
err分享
err收藏
OFF-eNET: An Optimally Fused Fully End-to-End Network for Automatic Dense Volumetric 3D Intracranial Blood Vessels Segmentation
err2020-01-01
err63
errOAAI
errNazir, Anam; Cheema, Muhammad Nadeem; Sheng, Bin; Li, Huating; Li, Ping; Yang, Po; Jung, Younhyun; Qin, Jing; Kim, Jinman; Feng, David Dagan
err分享
err收藏
Google Earth Shows Clandestine Worlds
err2010-08-27
err0
PREAI
errHeather Pringle
err分享
err收藏
学者 查看更多内容