Remote Sensing Scene Classification Based on Convolutional Neural Networks Pre-Trained Using Attention-Guided Sparse Filters

Remote Sensing

Jingbo Chen¹  Jiansheng Chen¹  Chengyi Wang¹  Dongxu He¹  Stephen Ackland²  Zhong Ma³ 

[1] Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences, Beijing 100094, China;School of Computer Science and Informatics, De Montfort University, Leicester LE1 9BH, UK;Xi’an Microelectronics Technology Institute, Xi’an 710065, China;

关键词: scene classification;    visual attention mechanism;    unsupervised learning;    sparse filters;    convolutional neural networks;   

DOI  :  10.3390/rs10020290

来源: DOAJ

【 摘 要 】

Semantic-level land-use scene classification is a challenging problem, in which deep learning methods, e.g., convolutional neural networks (CNNs), have shown remarkable capacity. However, a lack of sufficient labeled images has proved a hindrance to increasing the land-use scene classification accuracy of CNNs. Aiming at this problem, this paper proposes a CNN pre-training method under the guidance of a human visual attention mechanism. Specifically, a computational visual attention model is used to automatically extract salient regions in unlabeled images. Then, sparse filters are adopted to learn features from these salient regions, with the learnt parameters used to initialize the convolutional layers of the CNN. Finally, the CNN is further fine-tuned on labeled images. Experiments are performed on the UCMerced and AID datasets, which show that when combined with a demonstrative CNN, our method can achieve 2.24% higher accuracy than a plain CNN and can obtain an overall accuracy of 92.43% when combined with AlexNet. The results indicate that the proposed method can effectively improve CNN performance using easy-to-access unlabeled images and thus will enhance the performance of land-use scene classification especially when a large-scale labeled dataset is unavailable.

【 授权许可】

Unknown