Rancang bangun aplikasi monitoring karyawan departemen HRIS di meja kerja pada PT KYU

Fauzi, Mokhamad (2024) Rancang bangun aplikasi monitoring karyawan departemen HRIS di meja kerja pada PT KYU. Bachelor thesis, Universitas Pelita Harapan.

	Text (Title) Title.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (19kB)
	Text (Abstract) Abstract.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (197kB)
	Text (TOC) TOC.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (276kB)
	Text (Chapter 1) Chapter 1.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (202kB)
	Text (Chapter 2) Chapter 2.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (203kB)
	Text (Chapter 3) Chapter 3.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (213kB)
	Text (Chapter 4) Chapter 4.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (392kB)
	Text (Chapter 5) Chapter 5.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (405kB)
	Text (Chapter 6) Chapter 6.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (372kB)
	Text (Bibliography) Bibliography.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (293kB)
	Text (Appendices) Appendices.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (4MB)

Abstract

Skripsi ini membahas pengembangan perangkat yang dapat memonitor ketersediaan anggota HRIS di ruang kerja Departemen HRIS di PT KYU menggunakan tools pengenalan objek yaitu you only look once (YOLO) versi YOLOv3. Dalam mengenali objek YOLOv3 membutuhkan data training yang dibuat menggunakan dataset yang banyak agar mendapatkan data training yang baik. Penulis menggunakan data training Joseph Redmon yang merupakan data training dengan 80 class objek dengan salah satunya adalah objek person serta menggunakan data training yang dikumpulkan penulis menggunakan dataset gambar manusia yang diambil secara random dari internet sebanyak 105 gambar. Dari uji coba yang dilakukan didapatkan hasil bahwa menggunakan data training Joseph Redmon lebih baik dibandingkan data training yang dibuat penulis berdasarkan pendeteksian benda bulat yang menyerupai kepala manusia. Dengan data training Joseph Redmon dalam 60 detik benda bulat hanya terdeteksi sebanyak 11 kali dengan nilai confidence rata – rata 0.668355 dibadingkan data training penulis yang terdeteksi sebanyak 60 kali dengan nilai confidence rata – rata 0.820307. Hasil pendeteksian anggota HRIS pada ruang kerja dapat dilihat melalui aplikasi berbasis web dan dalam aplikasi tersebut karyawan di PT KYU dapat melakukan booking untuk bertemu dengan anggota HRIS. / This thesis discusses the development of a tool that can monitor the availability of HRIS staff members in the work space of HRIS Department of PT KYU using object recognition tools, namely you only look once (YOLO) version YOLOv3. Recognizing YOLOv3 objects requires training data to be created using a large dataset in order to get good training data. The author used Joseph Redmon's training data, which is training data with 80 object classes, one of which is a person object. The author also used training data using a dataset of 105 human images taken randomly from the internet. From the trials carried out, it was found that using Joseph Redmon's training data gave better results than the training data created by the author based on detecting non-human round objects that similar to human heads. With Joseph Redmon's training data, in 60 seconds round objects were only detected 11 times with an average confidence value of 0.668355 compared to the author's training data which was detected 60 times with an average confidence value of 0.820307. The results of detecting HRIS staff members in the work space can be observed via a web-based application and in this application employees at PT KYU can make bookings to meet HRIS staff members.

Actions (login required)

View Item