From the course: Computer Vision on the Raspberry Pi 4

Unlock the full course today

Join today to access over 22,700 courses taught by industry experts or purchase this course individually.

Creating CNNs with Keras

Creating CNNs with Keras

From the course: Computer Vision on the Raspberry Pi 4

Start my 1-month free trial

Creating CNNs with Keras

- In this video, we'll move away from OpenCV and look at a package called Keras. Unlike OpenCv, Keras focuses solely on neural networks and it provides a simple API for creating training and running many different types of neural networks. Because of its simplicity, Google hired Keras's creator and incorporated Keras into its TensorFlow package. Today, Keras is the default neural network API in TensorFlow, our usage of Keras will consist of four steps, the first is to create an instance of the Sequential class and to add layers that represent the layers of a neural network. The compile method defines the training process and the fit method, trains the network. Instead of executing the model, we're going to convert it to the TensorFlow Lite format. Then we'll save the model to a file that can be copied to the Raspberry Pi. Each neural network layer corresponds to a Keras class. The two convolution layers are represented by…

Contents