Depositar, Alvy L.; Francisco, Cheska Louisse A.; Zinampan, Joel S. Jr. 4 0
Artisan AI: An AI-enhanced crochet assistant / 6
6
Depositar, Alvy L.; Francisco, Cheska Louisse A.; Zinampan, Joel S. Jr.
-
-
- vii, 172 pp. 28 cm.
-
-
-
-
- .
- .
- 0 .
- .
- 0 .
Undergraduate Thesis: (Bachelor of Science in Information Technology) - Pamantasan ng Lungsod ng Maynila, 2024.
5
ABSTRACT: Crochet making is a hobby that is on a rise in popularity especially during the pandemic period where everyone was forced to stay within their households. Stitch counting is one of the most crucial steps in the process and is considered to be error-prone for chrochet artists. Mistakes with manual stitch counting often lead to miscounts, frustration, or even affect the project output. This study aims to develop a trained artificial intelligence model that can aid crochet artists in the stitch counting process. Based on studies of available convolutional neutral networks, Yolov8 computer vision model was selected for data training of crochet stitch images. It is included in a progressive web application that presents community engagement and creativity and its deployed website was tested by crochet artists from the Philippines. The website was successfully deployed using Heroku and ISO 25010:2011 software quality model is used to evaluate the progressive web application. Through the development of a progressive web application, the researchers have produced an automatic stitch counter that is powered by artificial intelligence, specifically the field of computer vision and object detection. The overall mean of the ISO Evaluation, computed by getting the average of all the mean scores of the individual criteria, is 4.50. This overall mean falls within the Satisfied range, affirming that the application meets the specified ISO standards and requirements. The results indicate that ArtisanAI, the created progressive web application, is beneficial for crochet stitch counting, invites creativity, and community engagement. The stitch counter of Artisan AI aids in the cumbersome task of manually counting the stitches in a crochet project and it reduces the chances of committing errors, avoids the possibilities of losing track of sticth count, and improves the overall efficiency of crochet making process. The crochet making process is expedited and made efficient by the developed crochet stitch counter. The results reveal that even the personaluzed feed through the recommendation features delights users because of their unique experience suited for their desires. An improved model version can be made through an expanded dataset and dedicated high-power computer for training.