RSCNet: An Efficient Remote Sensing Scene Classification Model Based on Lightweight Convolution Neural Networks

Abstract

This study aims at improving the efficiency of remote sensing scene classification (RSSC) through lightweight neural networks and to provide a possibility for large-scale, intelligent and real-time computation in performing RSSC for common devices. In this study, a lightweight RSSC model is proposed, which is named RSCNet. First, we use the lightweight ShuffleNet v2 network to extract the abstract features from the images, which can guarantee the efficiency of the model. Then, the weights of the backbone are initialized using transfer learning, allowing the model to learn by drawing on the knowledge of ImageNet. Second, to further improve the classification accuracy of the model, we propose to combine ShuffleNet v2 with an efficient channel attention mechanism that allows the features of the input classifier to be weighted. Third, we use a regularization technique during the training process, which utilizes label smoothing regularization to replace the original loss function. The experimental results show that the classification accuracy of RSCNet is 96.75% and 99.05% on the AID and UCMerced_LandUse datasets, respectively. The floating-point operations (FLOPs) of the proposed model are only 153.71 M, and the time spent for a single inference on the CPU is about 2.75 ms. Compared with existing RSSC methods, RSCNet achieves relatively high accuracy at a very small computational cost.

Efficient knowledge distillation for remote sensing image classification: a CNN-based approach

“…Compared to previous lightweight methods, our KDE-Net achieves excellent accuracy and maintains a compact volume. Furthermore, compared to previous NAS and KD (Chen et al, 2022a(Chen et al, , 2022b 1.3 None 96.24…”

“…Using transfer learning from ImageNet-1K to train a preexisting model provides a straightforward approach to constructing lightweight classifiers. For example, Yu et al (2020) and Chen et al (2022aChen et al ( , 2022b developed their classifiers using the lightweight MobileNet or ShuffleNet architectures. Similarly, Shi et al (2021) and Huang et al (2023) introduced their RSI classification methods, which involved custom-tuned CNN layers, attention modules or a ViT.…”

“…Incorporating functional modules into a pretrained deep model is a potential strategy to enhance the performance of the original model. In this domain, Alhichri et al (2021), Chen et al (2022aChen et al ( , 2022b and Zhao et al (2021) proposed analogous methods by integrating attention modules into a pre-trained CNN. Similarly, Shi et al (2022) and Bai et al (2022) introduced their methods using variants of standard convolution structures.…”

“…Typically, researchers have developed efficient models for RSI classification by either training a preexisting CNN with a compact capacity (Chen et al, 2022a(Chen et al, , 2022bYu et al, 2020) or using a custom-built CNN architecture combined with functional modules (Shi et al, 2021;Xu et al, 2022aXu et al, , 2022bXu et al, , 2022cXu et al, , 2022dXu et al, 2021). However, these methods often sacrifice generalization capability, resulting in below-average accuracy.…”

An Intelligent Waste-Sorting and Recycling Device Based on Improved EfficientNet

“…The convolutional neural network (CNN) is one of the main branches of deep learning, and it is the mainstream image-recognition method nowadays. With the rapid development of deep learning technology, CNN has made significant achievements in image classification [ 7 ]. Numerous researchers have used the CNN to solve waste image classification tasks [ 8 , 9 ], and have achieved a series of achievements.…”