Object Classification
In this section, you will learn about object detection and classification. The next step involves image classification for a dataset with more than two classes. The three different types of models for object classification we will cover are image classification, classification with localization, and detection:
- Image classification: This involves training with a set number of classes and then trying to determine which of those classes is shown in the image. Think of the MNIST handwriting dataset. For these problems, you'll use a traditional CNN.
- Classification with localization: With this type, the model tries to predict where the object is in the image space. For these models, you use a simplified You Only Look Once (YOLO) or R-CNN.
- Detection: The last type is detection. This is where your model can detect several different objects and where they are located. For this, you use YOLO or an R-CNN:
Figure 7.24: Object classification types...
