Barong Tagalog Design Classification Using YOLOV8

Along with most modern industries, the fashion industry has adopted the shift towards digital automation. The Barong Tagalog, an original Filipino garment dating back hundreds of years ago, is modernized by local artisans and remains a prominent item even in contemporary Filipino fashion. The aim of this study is to implement the YOLOv8 algorithm to create a model that can detect and classify four designs of Barong Tagalog namely: Art Deco, Ethnic, Special, and Traditional designs. The study makes use of the Raspberry Pi Model 4B and a Raspberry Pi Camera Module 2 to capture live images of the garments and classify them according to their respective design. The datasets used in the study are gathered from a local garment producer to train and validate the YOLOv8 model. The model is tested using 24 images for each classification, amounting to 96 total images, and resulted to an accuracy of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{8 4 \%}$</tex>. The results of this study may prove useful to the Barong Tagalog industry by improving their inventory management, as well as a more categorized catalogue for their customers.