Convolutional neural network based automatic object detection on aerial images
Letter: IEEE GRSL IF: 2.228
Target: Scene classification
- Using the model generated from one dataset to test another dataset.
- using pretrained VGG model by Caffe framework, except the final layer, because the classification target amount is not equal, to avoid overfitting. Just like the project we do in Kaggle, directly import trained VGG model. Its not about transfer learning methods
- Using Adagrad rather than SGD. Adagrad is an algorithm for gradient-based optimization that does just this: It adapts the learning rate to the parameters, performing larger updates for infrequent and smaller updates for frequent parameters. For this reason, it is well-suited for dealing with sparse data.
- based on cell, not on pixel
- the future work will also focus on multiple layers
- Mentioned that using a pretrained network proved to have better performance
- multiscale images input can improve the representation of size-varying objects
- often using UCM as a classification evaluation