Efficient CNN for high-resolution remote sensing imagery understanding.

The analysis of remote sensing data and the classification of images are complex problems because understanding spatial patterns and intricate geometric structures of the data from which meaningful interrelationship pixels are extracted is essential in the remote sensing community. CNN is one of the Machine Learning methods often used in Remote Sensing problems. However, high-resolution aerial view classification often leverages large-scale data with a huge number of parameters of the CNN model. A large number of parameters makes it hard to be applied to remote imaging peripherals because it requires high capacity of storage and memory. We propose a training framework to solve that problem that produces a model with minimum parameters without sacrificing accuracy. To this end, we train EfficientNet using Weighted Loss of INS to compensate imbalanced classes while keeping the upper layer frozen to preserve essential features deriving from transfer learning. Moreover, we utilize Sparse Regularization on the loss function to make the model focussing on the global object. Finally, as post-training, the trained model is pruned to reduce the number of parameters within the network. Using this method on the AID Dataset, the best results are achieved by EfficientNet-B0 with Freeze 2 Layer, INS Weighted Loss, Sparse regularization with λ = 0.001, and Global Unstructured Conv2D Pruning with an accuracy of 96.81% on test data with total parameters of 2,463,501. This study proves that Weighted Loss and Sparse Regularization can help the model to improve accuracy while Pruning enhances efficiency by reducing the number of parameters by half without significantly lowering performance. [ABSTRACT FROM AUTHOR]