In the second class of the training, we will teach you about Deep Learning and give you a better understanding and overview about the different Deep Learning architectures with a focus on Convolutional Neural Networks (CNN).
The lesson starts with a quick recap of Neural Networks and moves on to investigate the different problems that may occur when adding more and more hidden-layers. Then, we introduce some different types of hidden-layers, such as convolutional layers and pooling layers and talk about the benefits of these hidden-layers and when/how you should apply them. After that, we move on to the different Deep Learning architectures, like residual networks and inception modules.
We will also have a look at how we can use very large pre-trained Deep Learning models and customize them so they can be used in your projects with less training data. Lastly, we will show you how Deep Learning algorithms can be used to detect objects or recognize faces.
The training includes theory, demos and hands-on exercises.
After this training you will have gained knowledge about:
- The problems that may occur when you apply a deep Neural Network and how you can solve them
- Different types of hidden layers, like convolutional layers
- Different deep learning architectures, like residual networks
- How to use large pre-trained Deep Learning models for you own tasks with less data
- How Deep Learning models work that are able to recognize objects and faces
- Lab session to get hands-on experience in building your own Deep Learning models to recognize handwritten digits and reduce noise in images