In this chapter, we will cover the use of pre-trained neural networks from the MXNet Model Zoo and GitHub, compare network performance, and adjust them for solving custom computer vision problems.
The following topics will be covered in this chapter:
- Loading a pre-trained network from the MXNet Model Zoo
- Using Inception V3 and MobileNet v2 classifiers to predict image classes
- Using Inception V3 and MobileNet v2 to extract image features
- Using features extracted from Inception V3 to train a new model
- Applying the transfer learning technique and Inception V2 to solve custom problems