Asymmetric coordinate attention spectral-spatial feature fusion network for hyperspectral image classification

Cheng, Shuli; Wang, Liejun; Du, Anyu

doi:10.1038/s41598-021-97029-5

Cited by 12 publications

(6 citation statements)

References 76 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Coordinate Attention (CA) module is a kind of efficient attention module, which can be flexibly and conveniently embedded in various image classification models without increasing too much computing overhead [20].Its core idea is to decompose channel attention into two one-dimensional feature codes, and effectively integrate spatial coordinate information into the generated attention diagram. The advantage is that it can not only capture information across channels, but also capture accurate location information, which helps the model to locate features more accurately.The specific calculation process is shown in the figure.…”

Section: Attention Mechanismmentioning

confidence: 99%

Mobile-CANet: a new high-precision and lightweight method for rice pest classification

Wei,

Wang,

et al. 2024

International Conference on Computer Graphics, Artificial Intelligence, and Data Processing (ICCAID 2023)

View full text Add to dashboard Cite

Rice pests can adversely impact the quality and productivity of rice, and proper identification of pests is beneficial for growers to taking prompt steps to manage pests. Although the current crop pest identification model has achieved a high recognition accuracy in recent years with the application of deep learning, it is challenging to deploy on the edge devices with limited computing resources and low performance, and it cannot meet the needs of practical agricultural production. Therefore, we propose an effective real-time lightweight pest detection method, Mobile-CANet, which is based on the convolutional neural networks. This study develops a new network structure to enhance the feature extraction capabilities in order to address the issue of location information being lost during feature extraction. The capacity to extract features is enhanced by this network structure, which also incorporates transfer learning to enhance generalization ability. The results of the experiments show that this method's accuracy and number of parameters are superior to those of the existing picture classification models. On the publicly available dataset of rice pests, it can reach 85.3% accuracy with only 3.312M parameters, which can offer technical assistance for the implementation of mobile algorithms in the future.

show abstract

Section: Attention Mechanismmentioning

confidence: 99%

Mobile-CANet: a new high-precision and lightweight method for rice pest classification

Wei,

Wang,

et al. 2024

International Conference on Computer Graphics, Artificial Intelligence, and Data Processing (ICCAID 2023)

View full text Add to dashboard Cite

show abstract

“…The Coordinate Attention Block is a convolutional neural network module that attends to features selectively, based on their spatial locations in an image [39]. This is achieved by separately applying global pooling along the x and y axes to capture the mean and variance of the feature maps along each axis.…”

Section: E Coordinate Attention Mechanism Blockmentioning

confidence: 99%

GrapeLeafNet: A Dual-Track Feature Fusion Network With Inception-ResNet and Shuffle-Transformer for Accurate Grape Leaf Disease Identification

Karthik,

Menaka,

Ompirakash

et al. 2024

IEEE Access

View full text Add to dashboard Cite

Grapes are a widely cultivated crop in the horticultural industry, renowned for their unique flavor and nutritional benefits. However, this crop is highly susceptible to various diseases that can cause significant reductions in yield and quality, resulting in considerable financial losses. Therefore, it is imperative to identify these diseases to effectively manage their spread. Traditionally, the identification of grape leaf diseases has relied on scientific expertise and observational skills. However, with the advent of deep learning methods, it is now feasible to recognize disease patterns from images of infected leaves. In this research, we propose a novel dual-track feature fusion network titled 'GrapeLeafNet' for detecting grape leaf disease. It employs a dual-track feature fusion approach, combining Inception-ResNet blocks with CBAM for local feature extraction and Shuffle-Transformer for global feature extraction. The first track uses Inception-ResNet blocks to represent features at multiple scales and map significant features, and CBAM captures significant spatial and channel dependencies. The second track employs Shuffle-Transformer to extract long-term dependencies and complex global features in images. The extracted features are then fused using Coordinate attention, enabling the network to capture both local and global contextual information. Experimental results on the Grape leaf disease dataset from Plant Village demonstrate the effectiveness of the proposed network, achieving an accuracy of 99.56%. INDEX TERMSGrape leaf disease, deep learning, transformer, CNN, attention I. INTRODUCTION

show abstract

“…To sum up, the integration of attention modules with CNN models has experienced three stages, namely simple embedding, stage-wise fusion, and layer-wise fusion. In addition, attention modules can also be directly used for spatial-spectral feature extraction [32,33].…”

Section: Introductionmentioning

confidence: 99%

Low-Rank Constrained Attention-Enhanced Multiple Spatial–Spectral Feature Fusion for Small Sample Hyperspectral Image Classification

Feng¹,

Zhang²,

Zhang³

et al. 2023

Remote Sensing

View full text Add to dashboard Cite

Hyperspectral images contain rich features in both spectral and spatial domains, which bring opportunities for accurate recognition of similar materials and promote various fine-grained remote sensing applications. Although deep learning models have been extensively investigated in the field of hyperspectral image classification (HSIC) tasks, classification performance is still limited under small sample conditions, and this has been a longstanding problem. The features extracted by complex network structures with large model size are redundant to some extent and prone to overfitting. This paper proposes a low-rank constrained attention-enhanced multiple feature fusion network (LAMFN). Firstly, factor analysis is used to extract very few components that can describe the original data using covariance information to perform spectral feature preprocessing. Then, a lightweight attention-enhanced 3D convolution module is used for deep feature extraction, and the position-sensitive information is supplemented using a 2D coordinate attention. The above widely varying spatial–spectral feature groups are fused through a simple composite residual structure. Finally, low-rank second-order pooling is adopted to enhance the convolutional feature selectivity and achieve classification. Extensive experiments were conducted on four representative hyperspectral datasets with different spatial–spectral characteristics, namely Indian Pines (IP), Pavia Center (PC), Houston (HU), and WHU-HongHu (WHU). The contrast methods include several advanced models proposed recently, including residual CNNs, attention-based CNNs, and transformer-based models. Using only five samples per class for training, LAMFN achieved overall accuracies of 78.15%, 97.18%, 81.35%, and 87.93% on the above datasets, which has an improvement of 0.82%, 1.12%, 1.67%, and 0.89% compared to the second-best model. The running time of LAMFN is moderate. For example, the training time of LAMFN on the WHU dataset was 29.1 s, and the contrast models ranged from 3.0 s to 341.4 s. In addition, ablation experiments and comparisons with some advanced semi-supervised learning methods further validated the effectiveness of the proposed model designs.

show abstract

Asymmetric coordinate attention spectral-spatial feature fusion network for hyperspectral image classification

Cited by 12 publications

References 76 publications

Mobile-CANet: a new high-precision and lightweight method for rice pest classification

Mobile-CANet: a new high-precision and lightweight method for rice pest classification

GrapeLeafNet: A Dual-Track Feature Fusion Network With Inception-ResNet and Shuffle-Transformer for Accurate Grape Leaf Disease Identification

Low-Rank Constrained Attention-Enhanced Multiple Spatial–Spectral Feature Fusion for Small Sample Hyperspectral Image Classification

Contact Info

Product

Resources

About