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

Figure 11

Reconstruct features from the fifth layer of the pretrained model.