Abstract | ||
---|---|---|
In this letter, we investigate how to extract deep feature representations based on convolutional neural networks (CNN) for high-resolution remote-sensing imagery retrieval (HRRSIR). Two effective schemes are proposed to generate the final feature rep-resentations for similarity measure. In the first scheme, the deep features are extracted from the fully-connected and convolutional layers of the pre-trained CNN models, respectively; in the second scheme, we fine-tune the pre-trained CNN model using the target remote sensing dataset to learn dataset-specific features. The deep feature representations generated by the two schemes are evalu-ated on two public and challenging datasets. The experimental results indicate that the proposed schemes are able to achieve state-of-the-art performance due to the good transferability of the CNN models. |
Year | Venue | Field |
---|---|---|
2016 | arXiv: Computer Vision and Pattern Recognition | Computer vision,Pattern recognition,Computer science,Artificial intelligence |
DocType | Volume | Citations |
Journal | abs/1610.03023 | 1 |
PageRank | References | Authors |
0.35 | 26 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Weixun Zhou | 1 | 5 | 1.56 |
Congmin Li | 2 | 1 | 0.35 |