/benchmark-relu-selu

This project will observe 4 different choice of hyperparameter on their effect to CNN with ReLU and SELU as activation function

Primary LanguageJupyter Notebook

Benchmark of ReLU and SELU activation function

The Convolutional Neural Network (CNN) has shown an excellent performance in computer vision task. Applications of CNN include image classification, image semantic segmentation, object detection in images, etc.This project will observe 4 different choice of hyperparameter on their effect to CNN with ReLU and SELU as activation function

Network Confuguration

Training Accuracy on Cifar-10 using Xavier Initialization

Training Loss on SVHN with Adam Optimization

More detail report is in pdf file.