Pengembangan Aplikasi Liveness Detection Berbasis Android dan TensorFlow Lite

Mayzura, Wan Sabrina (2024) Pengembangan Aplikasi Liveness Detection Berbasis Android dan TensorFlow Lite. Project Report. [s.n.], [s.l.]. (Unpublished)

Text 5025211023-Project_Report.pdf - Accepted Version Restricted to Repository staff only Download (753kB) | Request a copy

Abstract

Perkembangan teknologi digital telah membawa inovasi signifikan dalam bidang keamanan, salah satunya adalah autentikasi biometrik berbasis pengenalan wajah. Namun, sistem ini masih rentan terhadap serangan spoofing, di mana pengguna dapat memalsukan identitas menggunakan foto atau video. Untuk mengatasi masalah ini, dikembangkan aplikasi Liveness Detection yang mampu mendeteksi gerakan kepala sebagai langkah validasi tambahan, memastikan bahwa pengguna yang terdeteksi adalah individu yang nyata. Aplikasi ini dirancang menggunakan library Jetpack Compose dari Kotlin dan arsitektur Three-layer Architecture diterapkan untuk memastikan modularitas pengembangan. Proses pengembangan aplikasi ini meliputi pengumpulan dataset dan pengintegrasian aplikasi android dengan model machine learning yang telah terlatih dalam bentuk Tensorflow Lite. Hasil pengujian menunjukkan bahwa aplikasi mampu mendeteksi gerakan kepala dengan tingkat akurasi tinggi, mendukung autentikasi biometrik yang lebih aman.
============================================================================================================================
The development of digital technology has brought significant innovations in the field of security, one of which is facial recognition-based biometric authentication. However, this system remains vulnerable to spoofing attacks, where users can falsify their identity using photos or videos. To address this issue, a Liveness Detection application was developed, capable of detecting head movements as an additional validation step, ensuring that the detected user is a real individual. This application is designed using the Jetpack Compose library from Kotlin, and the Three-layer Architecture is implemented to ensure development modularity. The application development process includes data collection and integrating the Android application with a pre-trained machine learning model in TensorFlow Lite format. Testing results show that the application can detect head movements with high accuracy, supporting more secure biometric authentication.

Item Type:	Monograph (Project Report)
Uncontrolled Keywords:	Autentikasi Biometrik, Liveness Detection, TensorFlow Lite, Kotlin, Jetpack Compose
Subjects:	T Technology > TA Engineering (General). Civil engineering (General) > TA1650 Face recognition. Optical pattern recognition.
Divisions:	Faculty of Information Technology > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Wan Sabrina Mayzura
Date Deposited:	27 Dec 2024 08:04
Last Modified:	27 Dec 2024 08:04
URI:	http://repository.its.ac.id/id/eprint/116055

Actions (login required)

View Item