Perbandingan Akurasi Arsitektur MobileNet dan EfficientNet dalam Mendeteksi Produk Kemasan
DOI:
https://doi.org/10.29407/g6txkh58Abstract
Penelitian ini membahas perbandingan performa dua arsitektur deep learning, yaitu MobileNet dan EfficientNet, dalam mendeteksi produk kemasan berbasis citra. Dataset yang digunakan terdiri dari 150 gambar tiga jenis produk kemasan, yaitu Le Minerale, Isoplus, dan Sprite, yang telah melalui proses anotasi dan dibagi menjadi data latih, validasi, dan uji. Proses pelatihan dilakukan selama 10 epoch dengan ukuran citra 448x448 piksel. Evaluasi dilakukan menggunakan metrik accuracy, precision, recall, dan f1-score. Hasil penelitian menunjukkan bahwa MobileNet memiliki performa sangat baik dengan akurasi 100% pada ketiga kelas produk, serta nilai precision, recall, dan f1-score sebesar 1.00 pada semua kategori. Sementara itu, EfficientNet menunjukkan hasil yang kurang optimal dengan akurasi hanya 33%, disertai bias prediksi terhadap satu kelas saja. Berdasarkan hasil tersebut, MobileNet direkomendasikan sebagai arsitektur yang lebih efisien dan andal dalam kasus deteksi produk kemasan berdasarkan dari hasil dan penelitian yang dilakukan.
Keywords:
Deep Learning, EfficientNet, MobileNet, Object Detection, Produk Kemasan##plugins.themes.default.displayStats.downloads##
References
[1] E. Sihabudin, R. Kumalasari Niswatin, and L. S. Wahyuniar, “Penerjemah Bahasa Isyarat Menggunakan Tensorflow,” Seminar Nasional Inovasi Teknologi, Jul. 2022.
[2] R. Fajri and F. Fitria, “KLIK: Kajian Ilmiah Informatika dan Komputer Pengembangan Real-Time Object Detection System pada Perangkat Single-Board Computer,” Media Online), vol. 4, no. 2, pp. 1154–1162, 2023, doi: 10.30865/klik.v4i2.1224.
[3] MAHFUZON AKHIAR, “PENERAPAN DEEP LEARNING MENGGUNAKAN CONVOLUTIONAL NEURAL NETWORK PADA KLASIFIKASI MOTIF BATIK,” November, Sep. 2021.
[4] Y. Miftahuddin and F. Zaelani, “Perbandingan Metode Efficientnet-B3 dan Mobilenet-V2 Untuk Identifikasi Jenis Buah-buahan Menggunakan Fitur Daun,” 2022.
[5] S. A. F. Harahap and I. Irmawan, “Performance comparison of MobileNet, EfficientNet, and Inception for predicting crop disease,” Sriwijaya Electrical and Computer Engineering Journal, vol. 1, no. 1, pp. 30–36, Feb. 2024, doi: 10.62420/selco.v1i1.4.
[6] Albertus Adrian Susanto, “IMPLEMENTASI DETEKSI OBJEK SECARA REAL-TIME,” Yogyakarta, Jul. 2023.
[7] B. Khasoggi, Ermatita, and Samsuryadi, “Efficient mobilenet architecture as image recognition on mobile and embedded devices,” Indonesian Journal of Electrical Engineering and Computer Science, vol. 16, no. 1, pp. 389–394, Oct. 2019, doi: 10.11591/ijeecs.v16.i1.pp389-394.
[8] T. Turahman, E. Hasmin, and K. Aryasa, “Analisis Perbandingan Metode Convolutional Neural Network (CNN) dan MobileNet dalam Klasifikasi Penyakit Daun Padi.” [Online]. Available: https://www.kaggle.com/datasets/vbookshelf/rice-
[9] M. Iqbal Burhanuddin, A. Syaifullah, S. Adeka Putra Jaya, M. Gabriel Somoal, and U. Muhammadiyah HAMKA, “Analisis Komparatif Model MobilenetV1 Dan EfficientnetB0 Dalam Klasifikasi Citra Empat Musim Menggunakan Transfer Learning,” 2025.
[10] Aryo Hardirega, Irsan Jaelani, and Minarto, “IMPLEMENTASI CONVOLUTIONAL NEURAL NETWORK (CNN) KLASIFIKASI MOTIF BATIK MENGGUNAKAN EFFICIENTNET-B1,” Kab. Purwakarta, Oct. 2024.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Lu'ay Baehaqie, Daniel Swanjaya, Danar Putra Pamungkas
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Copyright on any article is retained by the author(s).
- The author grants the journal, right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgment of the work’s authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal’s published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work.
- The article and any associated published material is distributed under the Creative Commons Attribution-ShareAlike 4.0 International License
