Hands-On Deep Learning for IoT

Convolutional neural networks

CNNs have achieved much and have been widely adopted in computer vision (for example, image recognition). In CNN networks, the connection schemes are significantly different compared to an MLP or DBN. A few of the convolutional layers are connected in a cascade style. Each layer is backed up by an ReLU layer, a pooling layer, additional convolutional layers (+ReLU), and another pooling layer, which is followed by a fully connected layer and a softmax layer. The following diagram is a schematic of the architecture of a CNN that's used for facial recognition, which takes facial images as input and predicts emotions such as anger, disgust, fear, happy, and sad:

A schematic architecture of a CNN used for facial ...