Saved in:
Bibliographic Details
Main Authors: Fu, Da, Rong, Mingfei, Kim, Eun-Hu, Huang, Hao, Pedrycz, Witold
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2405.04093
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914787089186816
author Fu, Da
Rong, Mingfei
Kim, Eun-Hu
Huang, Hao
Pedrycz, Witold
author_facet Fu, Da
Rong, Mingfei
Kim, Eun-Hu
Huang, Hao
Pedrycz, Witold
contents Accurate classification of fine-grained images remains a challenge in backbones based on convolutional operations or self-attention mechanisms. This study proposes novel dual-current neural networks (DCNN), which combine the advantages of convolutional operations and self-attention mechanisms to improve the accuracy of fine-grained image classification. The main novel design features for constructing a weakly supervised learning backbone model DCNN include (a) extracting heterogeneous data, (b) keeping the feature map resolution unchanged, (c) expanding the receptive field, and (d) fusing global representations and local features. Experimental results demonstrated that using DCNN as the backbone network for classifying certain fine-grained benchmark datasets achieved performance advantage improvements of 13.5--19.5% and 2.2--12.9%, respectively, compared to other advanced convolution or attention-based fine-grained backbones.
format Preprint
id arxiv_https___arxiv_org_abs_2405_04093
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle DCNN: Dual Cross-current Neural Networks Realized Using An Interactive Deep Learning Discriminator for Fine-grained Objects
Fu, Da
Rong, Mingfei
Kim, Eun-Hu
Huang, Hao
Pedrycz, Witold
Computer Vision and Pattern Recognition
Artificial Intelligence
Accurate classification of fine-grained images remains a challenge in backbones based on convolutional operations or self-attention mechanisms. This study proposes novel dual-current neural networks (DCNN), which combine the advantages of convolutional operations and self-attention mechanisms to improve the accuracy of fine-grained image classification. The main novel design features for constructing a weakly supervised learning backbone model DCNN include (a) extracting heterogeneous data, (b) keeping the feature map resolution unchanged, (c) expanding the receptive field, and (d) fusing global representations and local features. Experimental results demonstrated that using DCNN as the backbone network for classifying certain fine-grained benchmark datasets achieved performance advantage improvements of 13.5--19.5% and 2.2--12.9%, respectively, compared to other advanced convolution or attention-based fine-grained backbones.
title DCNN: Dual Cross-current Neural Networks Realized Using An Interactive Deep Learning Discriminator for Fine-grained Objects
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2405.04093