Desain Dan Pembuatan Avatar Untuk Digital Twin Konser Musik Menggunakan Metahuman Unreal Engine 5

Fisabilillah, Muhammad Daffa' (2025) Desain Dan Pembuatan Avatar Untuk Digital Twin Konser Musik Menggunakan Metahuman Unreal Engine 5. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5024211006-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (14MB) | Request a copy

Abstract

Penelitian ini bertujuan untuk mengembangkan digital twin musisi dalam bentuk avatar fotorealistik menggunakan teknologi Metahuman Unreal Engine 5. Avatar ini diintegrasikan dengan motion capture, face recognition, dan simulasi pakaian untuk menciptakan pengalaman konser virtual yang imersif. Metode yang digunakan meliputi pembuatan model 3D wajah dengan HyperHuman AI dan Polycam, desain pakaian menggunakan Marvelous Designer, serta implementasi gerakan tubuh dan ekspresi wajah melalui Move AI, Rokoko, eepMotion, Live Link Face dan Face Landmarker Link 02. Hasil penelitian menunjukkan bahwa avatar yang dihasilkan memiliki kemiripan visual yang tinggi, gerakan yang alami, dan ekspresi wajah yang realistis saat tampil di panggung virtual. Pengujian performa mencapai target minimal 30 FPS pada resolusi 1080p. Evaluasi lokasi blend shape mata dan mulut pada salah satu motion capture menggunakan Mediapipe membuktikan kelayakan teknologi ini untuk industri musik digital.
==================================================================================================================================
This research aims to develop a musician's digital twin in the form of a photorealistic avatar using Metahuman Unreal Engine 5 technology. The avatar is integrated with motion capture, face recognition, and clothing simulation to create an immersive virtual concert experience. The methods employed include creating 3D facial models using HyperHuman AI and Polycam, designing clothing with Marvelous Designer, and implementing body movements and facial expressions through Move AI, Rokoko, DeepMotion, Live Link Face and Face Landmarker Link 02. The research results demonstrate that the generated avatar achieves high visual similarity, natural movements, and realistic facial expressions when performing on a virtual stage. Performance testing reached the minimum target of 30 FPS at 1080p resolution. Evaluation of eye and mouth blend shape locations in one of the motion captures using Mediapipe proves the feasibility of this technology for the digital music industry.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Digital Twin, Konser Virtual, Metaverse, Metahuman, Face Recognition, Motion Capture
Subjects:	T Technology > TA Engineering (General). Civil engineering (General) > TA1650 Face recognition. Optical pattern recognition.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User:	Muhammad Daffa' Fisabilillah
Date Deposited:	28 Jul 2025 07:30
Last Modified:	28 Jul 2025 07:30
URI:	http://repository.its.ac.id/id/eprint/122160

Actions (login required)

View Item