Searching for CNN Architectures for Remote Sensing Scene Classification

Broni-Bediako, Clifford; Murata, Yuki; Mormille, Luiz H.; Atsumi, Masayasu

doi:10.1109/tgrs.2021.3097938

Cited by 28 publications

(17 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, the tradeoff is between the margin of 0.02% in classification accuracy and the extra computing power needed to manage an additional 3 million parameters. The SLGE-CNN architecture (Broni-Bediako et al, 2021) is 1.21% more accurate than our method. However, it took 9.6 GPU days for SLGE-CNN to achieve 600 training epochs compared to the 1h 15mins and 400 training cycles we used to train the framework proposed in this paper.…”

Section: Resultsmentioning

confidence: 75%

“…This availability, in turn, triggered an avalanche of applications in domains such as agriculture, environment monitoring, disaster risk analysis, climate change, urban development, surveillance, land mapping, and land use and land cover classification (LULC) (Bi et al, 2020;Z. Li et al, 2020;Balarabe and Jordanov, 2021;Broni-Bediako et al, 2021). Interestingly, most of these datasets used for the research and training scene classification systems have images with varying resolutions and dimensions.…”

Section: Introductionmentioning

confidence: 99%

“…For datasets with relatively larger samples, such as the AID dataset (Xia et al, 2017), the images are usually resized to a moderately smaller dimension without throwing away any vital information. In other datasets, the images are either fed into the CNNs models in their original form or resized to reduce the model training time (Broni-Bediako et al, 2021). Other datasets, notably the EuroSAT (Helber et al, 2019), have images with a tiny dimension of 64x64 pixels, which cannot fit into the first layer of some transfer learning models.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A Hybrid Dilation Approach for Remote Sensing Scene Image Classification

2022

IJCISIS

View full text Add to dashboard Cite

While fine-tuning a transfer learning model alleviates the need for a vast amount of training data, it still comes with a few challenges. One of them is the range of image dimensions that the input layer of a model accepts. This issue is of interest, especially in tasks that require the use of a transfer learning model. In scene classification, for instance, images could come in varying sizes that could be too large/small to be fed into the first layer of the architecture. While resizing could be used to trim images to a required shape, that is usually not possible for images with tiny dimensions, for example, in the case of the EuroSAT dataset. This paper proposes an Xception model-based framework that accepts images of arbitrary size and then resizes or interpolates them before extracting and enhancing the discriminative features using an adaptive dilation module. After applying the approach for scene classification problems and carrying out a number of experiments and simulations, we achieved 98.55% accuracy on the EuroSAT dataset, 99.22% on UCM, 96.15% on AID and 96.04% on the SIRI-WHU dataset, respectively. We also monitored the micro-average and macro-average ROC curve scores for all the datasets to further evaluate the proposed model's effectiveness.

show abstract

Section: Resultsmentioning

confidence: 75%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Hybrid Dilation Approach for Remote Sensing Scene Image Classification

2022

IJCISIS

View full text Add to dashboard Cite

show abstract

“…Therefore, the tradeoff is between the margin of 0.02% in classification accuracy and the extra computing power needed to manage an additional 3 million parameters. The SLGE-CNN architecture (Broni-Bediako et al, 2021) is 1.21% more accurate than our method. However, it took 9.6 GPU days for SLGE-CNN to achieve 600 training epochs compared to the 1h 15mins and 350 epochs it took to train the framework we proposed in this paper.…”

Section: Table 1 Performance Comparison Between Our Model and Other A...mentioning

confidence: 75%

Interpolation and Context Magnification Framework for Classification of Scene Images

2022

Proceedings of the 16th International Conference on Computer Graphics, Visualization, Computer Vision and Image Processing (CGV

View full text Add to dashboard Cite

Recently, there has been an upsurge in publicly available remote sensing image classification datasets. Standard CNNs and pre-trained architectures have been applied for scene classification tasks. However, transfer learning models accept specific image dimensions as the minimum required size for their respective image input layers. Depending on the size of the input image, the final feature map might not contain the discriminative information needed for accurately classifying the dataset categories. The proposed technique effectively enables and enhances a transfer learning model (Xception) to be applied to scene classification tasks. The model works on an adaptive framework that interpolates images and selects an appropriate dilation layer to enhance the quality of extracted features for improved classification. This approach is evaluated on the EuroSAT, a dataset with images of 64x64 pixels, UCM and AID datasets, respectively. We recorded 98.55%, 99.22%, and 96.15% accuracy for the EuroSAT, UCM, and AID datasets, respectively. Our model and the reported results have opened the potential of the Xception, which in our view, has not been given its fair share of attention, despite its efficient parameter utilisation.

show abstract

“…Recent semantic segmentation models frequently adopt Convolutional Neural Networks (CNNs), which have achieved significant advances in classical issues such as speech recognition, picture recognition, and natural language processing due to their robust feature learning and expression capabilities. Although it has been a research hotspot in remote sensing image interpretation and has yielded some achievements [25][26][27][28][29][30], the typical CNN model has significant limitations, including complex parameters and redundant calculations induced by the addition of relatively more layers.…”

Section: Introductionmentioning

confidence: 99%

Extraction of Rural Residential Land from Very-High Resolution UAV Images Using a Novel Semantic Segmentation Framework

et al. 2022

View full text Add to dashboard Cite

Accurate recognition and extraction of rural residential land (RRL) is significant for scientific planning, utilization, and management of rural land. Very-High Resolution (VHR) Unmanned Aerial Vehicle (UAV) images and deep learning techniques can provide data and methodological support for the target. However, RRL, as a complex land use assemblage, exhibits features of different scales under VHR images, as well as the presence of complex impervious layers and backgrounds such as natural surfaces and tree shadows in rural areas. It still needs further research to determine how to deal with multi-scale features and accurate edge features in such scenarios. In response to the above problems, a novel framework named cascaded dense dilated network (CDD-Net), which combines DenseNet, ASPP, and PointRend, is proposed for RRL extraction from VHR images. The advantages of the proposed framework are as follows: Firstly, DenseNet is used as a feature extraction network, allowing feature reuse and better network design with fewer parameters. Secondly, the ASPP module can better handle multi-scale features. Thirdly, PointRend is added to the model to improve the segmentation accuracy of the edges. The research takes a plain village in China as the research area. Experimental results show that the Precision, Recall, F1 score, and Dice coefficients of our approach are 91.41%, 93.86%, 92.62%, and 0.8359, respectively, higher than other advanced models used for comparison. It is feasible in the task of high-precision extraction of RRL using VHR UAV images. This research could provide technical support for rural land planning, analysis, and formulation of land management policies.

show abstract

Searching for CNN Architectures for Remote Sensing Scene Classification

Cited by 28 publications

References 57 publications

A Hybrid Dilation Approach for Remote Sensing Scene Image Classification

A Hybrid Dilation Approach for Remote Sensing Scene Image Classification

Interpolation and Context Magnification Framework for Classification of Scene Images

Extraction of Rural Residential Land from Very-High Resolution UAV Images Using a Novel Semantic Segmentation Framework

Contact Info

Product

Resources

About