The AlexNet, LeNet-5 and VGG NET applied to CIFAR-10
Xinche Zhang
Abstract
The problem this work trying to solve is applying AlexNet, LeNet-5 and VGG Net to the CIFAR-10 database and tell the features and performance of these structures. Multiple strategies of training have been tested, they are respectively AlexNet, LeNet5, VGG Net. These neural network algorithms are famous since they have achieved high accuracy in competitions. Although they are not designed for training CIFAR-10, it is worth to exploit that whether such an excellent model compatible to different input. Therefore, they will modified and tested the performance on CIFAR-10.
Topics & Concepts
Computer scienceExploitNet (polyhedron)Artificial intelligenceConvolutional neural networkArtificial neural networkPattern recognition (psychology)MathematicsGeometryComputer securityNeural Networks and ApplicationsAstronomical Observations and InstrumentationInfrared Target Detection Methodologies