TY - JOUR
T1 - ADM-Net
T2 - attentional-deconvolution module-based net for noise-coupled traffic sign recognition
AU - Chung, Jun Ho
AU - Kim, Dong Won
AU - Kang, Tae Koo
AU - Lim, Myo Taeg
N1 - Funding Information:
This work was funded by the Scientific and Technological Research Project of Henan Province (No. 202102110051 and 202102110054), and the Scientific Research Starting Foundation of Henan Institute of Science and Technology (No. 103010620001/003). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Funding Information:
The following grant information was disclosed by the authors: Scientific and Technological Research Project of Henan Province: 202102110051, 202102110054. Scientific Research Starting Foundation of Henan Institute of Science and Technology: 103010620001/003.
Publisher Copyright:
© 2022, The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.
PY - 2022/7
Y1 - 2022/7
N2 - Convolutional Neural Networks (CNNs) have become primary technologies in computer vision systems across multiple fields. Its central characteristic is to slide filters on input images and repeats the same procedures to obtain the image’s robust features. However, conventional CNNs struggle to classify objects when the input images are contaminated by unavoidable external noises such as missing information, blur, or illumination. This paper proposes an attentional-deconvolution module (ADM)-based net(ADM-Net) in which ADMs, convolutional-pooling, and a fully convolutional network (FCN) are applied to improve classification under such harsh conditions. The structure of ADM includes an attention layer, deconvolution layer and max-pooling. The attention layer and convolutional pooling help the proposed network maintain key features through convolution procedures under noise-coupled environments. The deconvolution layers and fully convolutional structure have advantages in providing additional information from upscale feature maps and enabling the network to store local pixel information. The ADM-Net was demonstrated on the German traffic sign recognition benchmark with different noise cases comparing densenet, multi-scale CNN, a committee of CNN, hierarchical CNN, and a multi-column deep neural network. Demonstrations of ADM-Net achieve the highest records in different cases such as 1) blur and missing information case: 86.637%, 2) missing information and illumination case: 92.329%, and 3) blur, missing information, and illumination case: 80.221%. Training datasets for ADM-Net have limited conditions, the proposed network demonstrates its robustness effectively under noise-coupled environments.
AB - Convolutional Neural Networks (CNNs) have become primary technologies in computer vision systems across multiple fields. Its central characteristic is to slide filters on input images and repeats the same procedures to obtain the image’s robust features. However, conventional CNNs struggle to classify objects when the input images are contaminated by unavoidable external noises such as missing information, blur, or illumination. This paper proposes an attentional-deconvolution module (ADM)-based net(ADM-Net) in which ADMs, convolutional-pooling, and a fully convolutional network (FCN) are applied to improve classification under such harsh conditions. The structure of ADM includes an attention layer, deconvolution layer and max-pooling. The attention layer and convolutional pooling help the proposed network maintain key features through convolution procedures under noise-coupled environments. The deconvolution layers and fully convolutional structure have advantages in providing additional information from upscale feature maps and enabling the network to store local pixel information. The ADM-Net was demonstrated on the German traffic sign recognition benchmark with different noise cases comparing densenet, multi-scale CNN, a committee of CNN, hierarchical CNN, and a multi-column deep neural network. Demonstrations of ADM-Net achieve the highest records in different cases such as 1) blur and missing information case: 86.637%, 2) missing information and illumination case: 92.329%, and 3) blur, missing information, and illumination case: 80.221%. Training datasets for ADM-Net have limited conditions, the proposed network demonstrates its robustness effectively under noise-coupled environments.
KW - Attention mechanism
KW - Convolutional neural network
KW - Deconvolution
KW - Fully convolutional network
KW - Traffic sign recognition
UR - http://www.scopus.com/inward/record.url?scp=85126534214&partnerID=8YFLogxK
U2 - 10.1007/s11042-022-12219-1
DO - 10.1007/s11042-022-12219-1
M3 - Article
AN - SCOPUS:85126534214
SN - 1380-7501
VL - 81
SP - 23373
EP - 23397
JO - Multimedia Tools and Applications
JF - Multimedia Tools and Applications
IS - 16
ER -