Skip to main content

Table 9 The parameter and hyperparameter settings of the model

From: Deep learning-driven multi-view multi-task image quality assessment method for chest CT image

Model

Image size

Optimizer

Initial learning rate

Loss function

YOLOv8

640 x 640

Stochastic gradient descent (SGD)

0.01

Binary cross entropy

U-Net

512 x 512

Adaptive moment estimation (ADAM)

0.0001

Cross entropy