Fauzan Kurnia Rahman, . (2025) PENERAPAN METODE CONVOLUTIONAL NEURAL NETWORK DALAM PLUGIN FIGMA UNTUK PEMERIKSAAN KESELARASAN KOMPONEN DESAIN DI PT BUKIT MAKMUR MANDIRI UTAMA. Skripsi thesis, Universitas Pembangunan Nasional Veteran Jakarta.
![[img]](http://repository.upnvj.ac.id/style/images/fileicons/text.png) |
Text
ABSTRAK.pdf Download (242kB) |
|
Text
AWAL.pdf Download (590kB) |
|
Text
BAB 1.pdf Restricted to Repository UPNVJ Only Download (320kB) |
|
Text
BAB 2.pdf Restricted to Repository UPNVJ Only Download (801kB) |
|
Text
BAB 3.pdf Restricted to Repository UPNVJ Only Download (549kB) |
|
Text
BAB 4.pdf Restricted to Repository UPNVJ Only Download (2MB) |
|
Text
BAB 5.pdf Download (284kB) |
|
Text
DAFTAR PUSTAKA.pdf Download (252kB) |
|
Text
DAFTAR RIWAYAT HIDUP.pdf Restricted to Repository UPNVJ Only Download (114kB) |
|
Text
LAMPIRAN.pdf Restricted to Repository UPNVJ Only Download (919kB) |
|
Text
HASIL PLAGIARISME.pdf Restricted to Repository staff only Download (17MB) |
|
Text
ARTIKEL KI.pdf Restricted to Repository staff only Download (693kB) |
Abstract
The process of verifying design component alignment in the PT Bukit Makmur Mandiri Utama Design System is still performed manually, resulting in time-consuming workflows and error-prone outcomes. This study aims to develop an automated Figma plugin based on a Convolutional Neural Network (CNN) to detect mismatched components and recommend replacements from the master Design System file. Development followed the Waterfall model, encompassing requirements identification, design, and implementation using Vue and TensorFlow.js. The dataset comprises 3,630 images across 121 component classes, split into 70% training, 15% validation, and 15% testing subsets, with data augmentation applied to prevent overfitting. The CNN model processes 224 × 224-pixel inputs and produces class predictions through convolutional, pooling, and fully connected layers. Initial evaluation showed an average validation accuracy above 0.88 and a top F1-score of 0.92 across 5 randomly selected classes. Functional testing via Black-Box Testing demonstrated that the plugin runs smoothly within Figma without any significant performance issues. A/B Testing further indicated a 46 % reduction in design checking time compared to manual methods. Additionally, User Acceptance Testing recorded a 96,44% user acceptance rate. These results indicate that the developed plugin significantly enhances the efficiency and accuracy of design component validation.
| Item Type: | Thesis (Skripsi) |
|---|---|
| Additional Information: | [No.Panggil: 2110511072] [Pembimbing 1: Musthofa Galih Pradana] [Pembimbing 2: Muhammad Panji Muslim] [Penguji 1: Neny Rosmawarni] [Penguji 2: I Wayan Rangga Pinastawa] |
| Uncontrolled Keywords: | Figma Plugin, Design System, Design Components, Convolutional Neural Network |
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software |
| Divisions: | Fakultas Ilmu Komputer > Program Studi Informatika (S1) |
| Depositing User: | FAUZAN KURNIA RAHMAN |
| Date Deposited: | 06 Aug 2025 06:49 |
| Last Modified: | 06 Aug 2025 06:49 |
| URI: | http://repository.upnvj.ac.id/id/eprint/37544 |
Actions (login required)
 |
View Item |