10
Q. Wang et al.
5 Conclusion
In this paper, we investigate the global feature descriptors extracted from the
last FC layer of CNN pre-trained models. Comparing to a single-model feature,
the fusion features are more representative. Both pre-trained CNN-based features, VGGNet-16 and Resnet-50, are proposed to form the last global feature
descriptors. The proposed method focuses on the most contribution features in
both different CNN models. The result of experiment illustrates that feature
fusion with a suitable dimension number of PCA transformation can enhance
the performance of classification. Comprehensive evaluations of the public RSI
scene classification perform the model training efficiency.
References
1. Bian X et al (2017) Fusing local and global features for high-resolution scene classification. IEEE J Sel Top Appl Earth Obs Remote Sens 10(6):2889–2901
2. Rublee E et al (2011) ORB: an efficient alternative to SIFT or SURF. In: 2011
IEEE international conference on computer vision (ICCV). IEEE
3. Salti S, Tombari F, Di Stefano L (2014) SHOT: unique signatures of histograms
for surface and texture description. Comput Vis Image Underst 125:251–264
4. Tombari, F, Salti S, Di Stefano L (2010) Unique signatures of histograms for local
surface description. In: European conference on computer vision. Springer, Berlin,
pp 356–369
5. Krizhevsky A, Sutskever I, Hinton GE (2012) Imagenet classification with deep
convolutional neural networks. In: Advances in neural information processing systems, pp 1097–1105
6. Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale
image recognition. arXiv preprint arXiv:1409.1556
7. He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition.
In: Proceedings of the IEEE conference on computer vision and pattern recognition,
pp 770–778
8. Liu N et al (2018) Exploiting convolutional neural networks with deeply local
description for remote sensing image classification. IEEE Access 6:11215–11228
9. Nogueira K, Penatti OAB, dos Santos JA (2017) Towards better exploiting convolutional neural networks for remote sensing scene classification. Pattern Recogn.
61:539–556
10. Kanter JM, Veeramachaneni K (2015) Deep feature synthesis: Towards automating
data science endeavors. In: 2015 IEEE international conference on data science and
advanced analytics (DSAA), 36678. IEEE, pp 1–10
11. Ke,Y, Sukthankar R (2004) PCA-SIFT: a more distinctive representation for local
image descriptors. In: Proceedings of the 2004 IEEE computer society conference
on computer vision and pattern recognition, 2004, vol 2. IEEE
12. Yang Y, Newsam S (2010) Bag-of-visual-words and spatial extensions for land-use
classification. In: Proceedings of the 18th SIGSPATIAL international conference
on advances in geographic information systems. ACM, pp 270–279
Q. Wang et al.
5 Conclusion
In this paper, we investigate the global feature descriptors extracted from the
last FC layer of CNN pre-trained models. Comparing to a single-model feature,
the fusion features are more representative. Both pre-trained CNN-based features, VGGNet-16 and Resnet-50, are proposed to form the last global feature
descriptors. The proposed method focuses on the most contribution features in
both different CNN models. The result of experiment illustrates that feature
fusion with a suitable dimension number of PCA transformation can enhance
the performance of classification. Comprehensive evaluations of the public RSI
scene classification perform the model training efficiency.
References
1. Bian X et al (2017) Fusing local and global features for high-resolution scene classification. IEEE J Sel Top Appl Earth Obs Remote Sens 10(6):2889–2901
2. Rublee E et al (2011) ORB: an efficient alternative to SIFT or SURF. In: 2011
IEEE international conference on computer vision (ICCV). IEEE
3. Salti S, Tombari F, Di Stefano L (2014) SHOT: unique signatures of histograms
for surface and texture description. Comput Vis Image Underst 125:251–264
4. Tombari, F, Salti S, Di Stefano L (2010) Unique signatures of histograms for local
surface description. In: European conference on computer vision. Springer, Berlin,
pp 356–369
5. Krizhevsky A, Sutskever I, Hinton GE (2012) Imagenet classification with deep
convolutional neural networks. In: Advances in neural information processing systems, pp 1097–1105
6. Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale
image recognition. arXiv preprint arXiv:1409.1556
7. He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition.
In: Proceedings of the IEEE conference on computer vision and pattern recognition,
pp 770–778
8. Liu N et al (2018) Exploiting convolutional neural networks with deeply local
description for remote sensing image classification. IEEE Access 6:11215–11228
9. Nogueira K, Penatti OAB, dos Santos JA (2017) Towards better exploiting convolutional neural networks for remote sensing scene classification. Pattern Recogn.
61:539–556
10. Kanter JM, Veeramachaneni K (2015) Deep feature synthesis: Towards automating
data science endeavors. In: 2015 IEEE international conference on data science and
advanced analytics (DSAA), 36678. IEEE, pp 1–10
11. Ke,Y, Sukthankar R (2004) PCA-SIFT: a more distinctive representation for local
image descriptors. In: Proceedings of the 2004 IEEE computer society conference
on computer vision and pattern recognition, 2004, vol 2. IEEE
12. Yang Y, Newsam S (2010) Bag-of-visual-words and spatial extensions for land-use
classification. In: Proceedings of the 18th SIGSPATIAL international conference
on advances in geographic information systems. ACM, pp 270–279