Bag of Visual Words Model with Deep Spatial Features for Geographical Scene Classification

Figure 10

Reconstruct features from the third layer of the pretrained model.