Classification of NTT Woven Motif Images Based on Convolutional Neural Networks

This research aims to develop an image classification model for East Nusa Tenggara (NTT) weaving motifs using Convolutional Neural Network (CNN). Data augmentation techniques are applied to increase the generalization ability of the model in overcoming dataset limitations. The NTT weaving image dataset consists of six motif classes which are used for model training and testing. Evaluation of model performance is carried out using accuracy, precision, recall and F1-score metrics. The research results show that the CNN model trained for 100 epochs achieved $\mathbf{9 3. 5 2 \%}$ accuracy, $\mathbf{9 3. 5 1 \%}$ precision, $\mathbf{9 3. 7 9 \%}$ recall, and $93.65 \%$ F1-score. Increasing the number of epochs proved effective in improving model performance, although computing time also increased. In conclusion, CNN is able to provide an efficient solution for classifying NTT weaving motifs with high accuracy, and this research can support cultural preservation and the development of technology-based creative industries.