KSC 2019
Current Result Document :
ÇѱÛÁ¦¸ñ(Korean Title) |
ÇÏÀÌÆÛÆĶó¹ÌÅÍ ÃÖÀûȸ¦ ÅëÇÑ µö ÄÁº¼·ç¼Ç ½Å°æ¸ÁÀÇ ¼º´É Çâ»ó |
¿µ¹®Á¦¸ñ(English Title) |
Performance Improvements of Deep Residual Convolutional Network with Hyperparameter Optimizations |
ÀúÀÚ(Author) |
¾Æ¹Ù½º ÀÚÆĸ£
À̸íÈ£
Abbas Jafar
Lee Myungho
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 46 NO. 02 PP. 0013 ~ 0015 (2019. 12) |
Çѱ۳»¿ë (Korean Abstract) |
|
¿µ¹®³»¿ë (English Abstract) |
Convolutional Neural Networks (CNNs) is one of the most commonly used deep learning models to train a large number of datasets and getting valuable results in image recognition. Deep Residual Learning (ResNet) is one of the most famous CNN for the computer vision tasks that won the ILSVR-2015 classification competition. ResNet is also one of the deepest models to train the neural networks with the idea of identity mapping for short connections. In this paper, the classification error rate of the ResNet model for the CIFAR-10 dataset is improved by optimizing hyperparameters. Our method improves performance by considering the computational complexity of the model.
|
Å°¿öµå(Keyword) |
Image recognition
Convolutional Neural Network (CNN)
Deep Residual Learning
Hyperparameters
Stochastic Gradient Descent
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|