Identification of batik making method from images using convolutional neural network with limited amount of data

This study aims to apply the convolutional neural network (CNN) classify batik based on its manufacturing method, namely Batik Tulis which are hand drawn, Cap where stamps used create pattern, and Printing printed using textile printing machine. We collected 40 images for each type of with a total 120 images. To speed up simplify model building process, we implemented transfer learning 3 basic CNN architectures, ResNet, DenseNet, VGG batch normalization. also experimented new dataset by breaking image down into 30 smaller Image augmentation was prevent overfitting as well provide variations in training data. The experimental results 5-fold cross validation show that densenet169 gives best original an accuracy 79.17% while vgg13_bn shows performance modified 87.61%. All models showed increase when dataset, except did not significant difference performance.