Smart Learning Library

  • Beranda
  • Informasi
  • News
  • Bantuan
  • Pustakawan
  • Area Anggota
  • Pilih Bahasa :
    Arabic Bengali Brazilian Portuguese English Espanol German Indonesia Bahasa Jepang Melayu Persia Russian Thai Turkish Urdu

Search by:

All Author Subject ISBN/ISSN Advanced Search

Last search:

{{tmpObj[k].text}}
Image of KLASIFIKASI CITRA JENIS KULIT MENGGUNAKAN CONVOLUTIONAL NEURAL NETWORK (CNN) EFFICIENTNET-B0 DAN PYTORCH UNTUK IMPLEMENTASI DI APLIKASI MOBILE

SKRIPSI IF

KLASIFIKASI CITRA JENIS KULIT MENGGUNAKAN CONVOLUTIONAL NEURAL NETWORK (CNN) EFFICIENTNET-B0 DAN PYTORCH UNTUK IMPLEMENTASI DI APLIKASI MOBILE

PRATAMA, ANDREAS KEVIN - Personal Name;

ABSTRAK

Penelitian ini membahas pengembangan sistem klasifikasi jenis kulit manusia berbasis citra wajah menggunakan Convolutional Neural Network (CNN) dengan arsitektur EfficientNet-B0, yang diimplementasikan ke dalam aplikasi mobile berbasis Android menggunakan PyTorch Mobile. Tujuan penelitian adalah merancang model yang mampu mengklasifikasikan tiga jenis kulit—kering (dry), normal, dan berminyak (oily)—secara otomatis, cepat, dan akurat.

Dataset yang digunakan terdiri dari citra wajah berlabel jenis kulit yang dibagi menjadi data pelatihan, validasi, dan pengujian menggunakan metode stratified split. Pra-pemrosesan dilakukan melalui resize, normalisasi, dan augmentasi citra (rotasi acak, horizontal flip). Model EfficientNet-B0 dimodifikasi pada layer output menjadi tiga neuron sesuai jumlah kelas, menggunakan bobot awal dari ImageNet, dan dilatih dengan fine-tuning seluruh layer. Fungsi loss yang digunakan adalah CrossEntropyLoss dengan bobot kelas (weighted loss) untuk mengatasi ketidakseimbangan data.

Hasil pengujian menunjukkan bahwa model mencapai akurasi validasi tertinggi sebesar 94,62% dengan macro average F1-score sebesar 0,95. Confusion matrix menunjukkan tingkat kesalahan prediksi yang rendah di seluruh kelas, meskipun performa pada kelas “dry” sedikit lebih rendah dibanding kelas lain. Evaluasi oleh tenaga medis menunjukkan bahwa meskipun sistem memiliki akurasi tinggi pada dataset, penggunaannya dalam praktik medis masih memerlukan data asli dari pasien serta observasi langsung.

Kata kunci: Klasifikasi jenis kulit, CNN, EfficientNet-B0, PyTorch Mobile, Android.


ABSTRACT

This research presents the development of a human skin type classification system based on facial images using a Convolutional Neural Network (CNN) with the EfficientNet-B0 architecture, implemented into an Android-based mobile application via PyTorch Mobile. The objective is to design a model capable of classifying three skin types—dry, normal, and oily—automatically, quickly, and accurately.

The dataset consists of labeled facial images, split into training, validation, and testing sets using a stratified split method. Preprocessing includes resizing, normalization, and image augmentation (random rotation, horizontal flip). The EfficientNet-B0 model was modified in the output layer to produce three neurons corresponding to the number of classes, initialized with ImageNet weights, and trained with full-layer fine-tuning. CrossEntropyLoss with class weighting (weighted loss) was applied to address class imbalance.

Experimental results show the model achieved a highest validation accuracy of 94.62% with a macro average F1-score of 0.95. The confusion matrix indicates low misclassification rates across all classes, although performance for the “dry” class was slightly lower than the others. Medical evaluation revealed that despite the high accuracy on the dataset, real-world medical usage still requires authentic patient data and direct observation.

Keywords: Skin type classification, CNN, EfficientNet-B0, PyTorch Mobile, Android.


Ketersediaan
S250927005607.2 PRA kPerpustakaan STMIK AMIKBANDUNGTersedia
Informasi Detil
Judul Seri
-
No. Panggil
607.2 PRA k
Penerbit
: ., 2025
Deskripsi Fisik
-
Bahasa
Indonesia
ISBN/ISSN
-
Klasifikasi
NONE
Tipe Isi
-
Tipe Media
-
Tipe Pembawa
-
Edisi
-
Subyek
-
Info Detil Spesifik
-
Pernyataan Tanggungjawab
-
Versi lain/terkait

Tidak tersedia versi lain

Lampiran Berkas
  • JURNAL
Komentar

You must be logged in to post a comment

Smart Learning Library
  • Information
  • Services
  • Librarian
  • Member Area

About Us

As a complete Library Management System, SLiMS (Senayan Library Management System) has many features that will help libraries and librarians to do their job easily and quickly. Follow this link to show some features provided by SLiMS.

Search

start it by typing one or more keywords for title, author or subject

Keep SLiMS Alive Want to Contribute?

© 2026 — Senayan Developer Community

Powered by SLiMS
Select the topic you are interested in
  • Computer science, information & general works
  • Philosophy & psychology
  • Religion
  • Social sciences
  • Language
  • Pure Science
  • Applied sciences
  • Arts & recreation
  • Literature
  • History & geography
Advanced Search