Team Members: Emre Tunç, Muhammed Sezer Berker, Tolga Furkan Güler
Hello everyone, this is the next blog of our White Blood Cell Classification project. This week, we will share with you which model to use and which approaches we intend to use to train the data set we have. Since the data set contained many image, we chose the Convolutional Neural Network (CNN) algorithm during the training process.
What is CNN?
Convolutional neural networks are used in fields such as object detection, image classification and image segmentation through images and video and are composed of layers of many different functions.
Convolutional Layer: The features of the picture are extracted with the help of the filters applied in this layer. In order to remove the high level features in the picture, the characteristics of the applied filter have an important place.
Pooling Layer:This is to reduce the number of property parameters in the network to reduce the number of computations that are applied to the layer. The most popular applied pooling methods are Max Pooling and Average Pooling.
Fully-Connected Layer: In this layer, learning process takes place. After the features of the picture are removed, we apply the flattening process to the existing matrix and learn through the neural network.
Output Layer:With the help of the classification algorithms used in the last layer, it is found out which class the picture belongs to. Softmax is one of the algorithms used.
We examined the related works and used models like Vgg16 and ResNet to use in the training process.
Thank you for reading and for your time.See you next week.
