Performasi Deteksi Jumlah Manusia Menggunakan YOLOv8
DOI:
https://doi.org/10.26905/jasiek.v5i2.11605Keywords:
Kecerdasan buatan, Jumlah kepala, YOLOV8, Roboflow, Confusion MatrixAbstract
Pengembangan deteksi kepala sudah meningkat dengan adanya peningkatan algoritma kecerdasan buatan. Peningkatan ini dapat pula dengan penambahan tugas yaitu menghitung jumlah orang dengan mendeteksi jumlah kepala. Tujuan penelitian ini adalah menentukan performansi model sistem penghitung jumlah kepala dengan menggunakan algoritma Yolov8. Penelitian ini hanya berfokus membuat model deteksi jumlah orang. Jumlah dataset yang dirancang berjumlah 2390 gambar yang diperoleh dari dataset Roboflow, dengan pemisahan data sebesar 70:20:10 untuk masing-masing, data latih; data uji ; data validasi. Besar Epoch pada pelatihan model yang digunakan adalah 50. Algoritma deteksi jumlah kepala meggunakan YOLOv8. Nilai yang diukur adalah performasi dari model data training, nilai confusion matrix dan nilai evaluasi dari confusion matrix. Nilai evaluasi yang akan dihitung adalah nilai presisi, nilai akurasi, recall dan f1-score.  Diperoleh hasil pengujian nilai akurasi sebesar 87,56 %, nilai presisi 83,74%, nilai recall 100% dan nilai F1-score 91,15%. Kurva presisi memberikan nilai tertinggi 1 pada tingkat kepercayaan 0,857, recall bernilai 0,8 pada tingkat kepercayaan 0, f1 0,716 pada kepercayaan 0,36 dan presisi-recall 0,771 pada 0,5 mAP. Berdasarkan nilai ini, model sudah cukup mendeteksi jumlah kepala.
References
Malhotra P, Garg E. Object detection techniques: A comparison. 2020 7th Int Conf Smart Struct Syst ICSSS 2020 2020:4–7. https://doi.org/10.1109/ICSSS49621.2020.9202254.
Motwani NP, S S. Human Activities Detection using DeepLearning Technique- YOLOv8. ITM Web Conf 2023;56:03003. https://doi.org/10.1051/itmconf/20235603003.
Wang G, Chen Y, An P, Hong H, Hu J, Huang T. UAV-YOLOv8: A Small-Object-Detection Model Based on Improved YOLOv8 for UAV Aerial Photography Scenarios. Sensors 2023;23. https://doi.org/10.3390/s23167190.
Yang G, Wang J, Nie Z, Yang H, Yu S. A Lightweight YOLOv8 Tomato Detection Algorithm Combining Feature Enhancement and Attention 2023.
Aboah A, Wang B, Bagci U, Adu-Gyamfi Y. Real-time Multi-Class Helmet Violation Detection Using Few-Shot Data Sampling Technique and YOLOv8. IEEE Comput Soc Conf Comput Vis Pattern Recognit Work 2023;2023-June:5350–8. https://doi.org/10.1109/CVPRW59228.2023.00564.
Wang X, Gao H, Jia Z, Li Z. BL-YOLOv8: An Improved Road Defect Detection Model Based on YOLOv8. Sensors (Basel) 2023;23. https://doi.org/10.3390/s23208361.
Bai R, Shen F, Wang M, Lu J, Zhang Z. Improving Detection Capabilities of YOLOv8-n for Small Objects in Remote Sensing Imagery: Towards Better Precision with Simpliied Model Complexity Improving Detection Capabilities of YOLOv8-n for Small Objects in Remote Sensing Imagery: Towards Better Pre. Res Sq 2023:0–9.
Sary IP, Andromeda S, Armin EU. Performance Comparison of YOLOv5 and YOLOv8 Architectures in Human Detection using Aerial Images. Ultim Comput J Sist Komput 2023;15:8–13. https://doi.org/10.31937/sk.v15i1.3204.
Diwan T, Anirudh G, Tembhurne J V. Object detection using YOLO: challenges, architectural successors, datasets and applications. Multimed Tools Appl 2023;82:9243–75. https://doi.org/10.1007/s11042-022-13644-y.
Downloads
Published
Issue
Section
License
Authors who publish with this journal agree to the following terms:
The journal allow the authors to hold the copyright without restrictions and allow the authors to retain publishing rights without restrictions.
Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution-ShareAlike 4.0 International License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
