Another simpler but usually less effective way of doing TL is to use a network trained on a specific task as a feature extractor. In this way, the feature we will extract will be very dependent on the task.
But we also know that the features created in different layers follow a hierarchical structure that will learn a high-level representation of the image in the following different layers:
- Lower layer: Features in lower layers will be very low-level. This means that they are quite generic and simple. Examples of features extracted in the first layer can be lines, edges, or linear relationships; we saw previously that, with one layer, we can describe linear relationships. The second layer will be able to capture more complex shapes, such as curves.
- Higher layer: Features in higher layers will be more high-level descriptions of our inputs. Parts of it might...