Skripsi

SISTEM DETEKSI OBJEK, PENGENALAN WAJAH, DAN PENGENALAN EKSPRESI WAJAH MENGGUNAKAN METODE MODIFIKASI ALEXNET

---

Penilaian

---

Service robots are one of the implementations of Industry 4.0 technology that require a computer vision system to detect objects, recognize faces, and identify facial expressions in real-time. However, most previous studies still separated these three functions into different systems, resulting in less efficient and less interactive robot performance. This study developed an integrated system based on deep learning by modifying the AlexNet architecture to enable simultaneous object detection, face recognition, and facial expression recognition. Image processing was performed using the Python programming language and tested directly on a service robot at the Control and Robotics Laboratory, Universitas Sriwijaya. Object detection was carried out using YOLOv8, while face and expression recognition were performed using the Modified AlexNet. Facial expressions were classified into five categories: happy, sad, angry, normal, and shocked. Based on the test results, the Modified AlexNet achieved a face detection accuracy of 78% and expression recognition accuracy of 74% after 50 epochs of training, significantly outperforming the standard AlexNet, which only achieved 32% and 40%, respectively. In real-time testing, the Modified AlexNet achieved 80% accuracy for face detection and 93% for expression recognition, with a distance estimation error of approximately ±1 cm. YOLOv8 demonstrated the highest accuracy in object detection at 82%, while Faster R-CNN showed poor performance with only 8% accuracy and failed to detect faces and expressions. The results indicate that the combination of YOLOv8 and Modified AlexNet offers an optimal and reliable solution to support intelligent and responsive service robot interactions.

---

Availability

Inventory Code	Barcode	Call Number	Location	Status
2507005591	T183724	T1837242025	Central Library (Reference)	Available but not for loan - Not for Loan

Detail Information

Series Title

-

Call Number

T1837242025

Publisher

Indralaya : Prodi Teknik Elektro, Fakultas Teknik Universitas Sriwijaya., 2025

Collation

xvi, 109 hlm.; ilus,; tab, 29 cm

Language

Indonesia

ISBN/ISSN

-

Classification

621.382 07

Content Type

Text

Media Type

unmediated

Carrier Type

other (computer)

Edition

-

Subject(s)

Teknik Komunikasi
Prodi Teknik Elektro

Specific Detail Info

-

Statement of Responsibility

SEPTA

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment