Klasifikasi Pakaian Berdasarkan Gambar Menggunakan Metode YOLOv3 dan CNN

Authors

  • Michael Christianto Wujaya Program Studi Informatika
  • Leo Willyanto Santoso Program Studi Informatika

Keywords:

Hotel resor, Arsitektur Neo – vernakular, Penginapan, Wisatawan, Pariwisata.

Abstract

Clothing is one of the primary human needs and have many functions. It’s function not solely to cover and protect the wearer, but also to look stylish. Mass media, the internet, and social media are the main place for people to find inspiration to look fashionable. But sometimes it is difficult to determine the type of clothing so it will be easy to find. Therefore, a program that is able to differentiate and classify clothes will be a great help.

The method we used are You Only Look Once to detect the clothing object from an image. The output of detection will be cropped and the result will be processed and classified by Convolutional Neural Network using ResNet50 architecture. In the training process of ResNet50, various things will be tuned which is learning rate, dropout, epoch, number of dense layer and its value, freezing layer, and data augmentation. Then program will search similar image using k-nearest neighbor.

The result of this study will classify clothes in an image that is worn by the model in the image. The average accuracy obtained using the fine-tuned ResNet50 is 86.44%.

References

[1] He, K,. Zhang, X., Ren, S., & Sun, J. 2015. Deep Residual Learning for Image Recognition.

[2] Jatmiko, B. P. 2019. Fashion Jadi Produk Terlaris dan Menguntungkan bagi E-Commerce. Kompas. Retrieved from https://money.kompas.com/read/2019/07/12/193100126/fashion-jadi-produk-terlaris-dan-menguntungkan-bagi-e-commerce

[3] Liu, R., Yan, Z., Wang, Z., & Ding, S., 2019. An Improved YOLOV3 for Pedestrian Clothing Detection, 2019 6th International Conference on Systems and Informatics (ICSAI), Shanghai, China, 2019, pp. 139-143, doi: 10.1109/ICSAI48974.2019.9010512.

[4] Mahadhirka, W. A. 2020. Ini Kategori Produk yang Paling Dicari di E-Commerce 2019. Kompas. Retrieved from https://money.kompas.com/read/2020/01/09/165909226/ini-kategori-produk-yang-paling-dicari-di-e-commerce-2019

[5] Prabhu. 2018. Understanding of Convolutional Neural Network (CNN) — Deep Learning. Retrieved from https://medium.com/@RaghavPrabhu/understanding-of-convolutional-neural-network-cnn-deep-learning-99760835f148

[6] Ramadhani, R. D. 2019. Memahami K-Nearest Neighbor (KNN) Dengan R . Retrieved from https://medium.com/@16611129/memahami-k-nearest-neighbor-knn-dengan-r-de5280439053

[7] Redmon, J., Divvala, S., Girshick, R., & Farhadi, A. 2016. You Only Look Once : Unified, Real-Time Object Detection

[8] Redmon, J., & Farhadi, A. 2018. YOLOv3: An Incremental Improvement

[9] Widiarini, A. D. 2019. Terus Tumbuh Signifikan, Begini Masa Depan E-Commerce di Indonesia. Kompas. Retrieved from https://money.kompas.com/read/2019/12/10/110500326/terus-tumbuh-signifikan-begini-masa-depan-e-commerce-di-indonesia

Downloads

Published

2021-04-10

Issue

Section

Articles