Claim Missing Document
Check
Articles

Found 2 Documents
Search

IMPLEMENTASI ALGORITMA KLASIFIKASI GERAKAN TANGAN POPULER MENGGUNAKAN CUSTOM MODEL CONVOLUTIONAL NEURAL NETWORK (CNN) Rizmawan, Dilan; Latifa, Ulinnuha
Jurnal Teknik Elektro Vol 14 No 2 (2024): Jurnal Teknik Elektro
Publisher : Program Studi Teknik Elektro Fakultas Teknik Universitas Palembang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.36546/jte.v14i2.1445

Abstract

Pada penelitian ini, diimplementasikan Convolutional Neural Network (CNN) dengan model kustom untuk mengklasifikasikan Gerakan tangan populer. Dataset dibuat sendiri dengan contoh gerakan tangan yang diambil. Pengujian dilakukan dengan 361 epoch dan batch size 256. Hasil menunjukkan peningkatan konsisten dalam performa model, meskipun akurasi validasi tidak mencapai lebih dari 0.94. Evaluasi model mencapai akurasi 94% dengan nilai loss yang rendah. Laporan klasifikasi menunjukkan kekurangan pada beberapa kelas seperti tumbs down, peace, stop be strong shake sign, bang-bang, dan rock. Namun, kelas seperti tumbs up dan ok memiliki performa terbaik dengan nilai rata-rata 1.00. Dalam penelitian ini, juga menggunakan metode Transfer Learning untuk memperbaiki performa model CNN.
Robot Autonomos Penyortiran dan Pemindahan Barang (Ramos) Wijaya, Andre; Rizmawan, Dilan; Fahrezi, Ari; Firdaus, Efrans; Oriza Syahda, Rizky; Jati Rochman, Wisnu Kuncoro; Hidayat, Rahmat
Jurnal Komputer dan Elektro Sains Vol. 3 No. 1 (2025): komets
Publisher : Sultan Publisher

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.58291/komets.v3i1.156

Abstract

In the current era, as we know, the industry in the field of shipping or logistics services has experienced very rapid growth in Indonesia. Logistics is a term for storing goods on a large scale which cannot be separated from the industrial sector, including the logistics delivery services industry. The development of the expedition services market in Indonesia was stated by Mr. Feriadi, General Chair of the DPP of the Association of Indonesian Express, Postal and Logistics Delivery Service Companies 4 (Asperindo), one of which was caused by the e-commerce market which has developed rapidly. Seeing this problem, we offer a solution, namely a tool that can steer and move goods according to the destination point that has been determined based on the goods code. The product to be designed is named RAMOS (Robot Autonomous Sorter and Mover). This robot can be used in a warehouse and the robot can walk autonomously according to a predetermined destination point. Functions as a tool for sorting and moving goods in logistics warehouses. This product can reduce human effort in moving goods that have been classified, and can reduce errors in the placement of goods that have been determined. RAMOS (Autonomous Robot Sorting and Moving Goods). This robot can be used in a warehouse and the robot can walk autonomously according to a predetermined destination point. Functions as a tool for sorting and moving goods in logistics warehouses. This product can reduce human effort in moving goods that have been classified, and can reduce errors in the placement of goods that have been determined. Then, there is an advantage to this robot, namely that it is able to avoid obstacles or collisions.