Skripsi
PENDETEKSIAN WAJAH BERKELOMPOK MENGGUNAKAN DEEP LEARNING
Face detection is one of the main challenges in the field of computer vision and artificial intelligence. The development of this particular technology has been highlighted due to its wide range of applications such as security surveillance systems, crowd analysis systems, and can be collaborated with face recognition systems for face identification purposes. The challenge emerges when attempting to detect a significant number of faces with various dimensions. Therefore, this research introduces the RetinaNet method, a deep learning-based approach for face detection that effectively identifies numerous individual faces or groups of faces. The grouped face detection system is built with a pre-trained model in order to produce the model weights. The backbone of the RetinaNet model used ResNet50 and ResNet101. The tests were conducted using 100 testing data points, achieving the highest Average Precision (AP) of 92.3%. The model is capable to detect facial objects across various scales, achieving APs of 45.3% for small objects, 62.9% for medium-sized objects, and 73.5% for large objects.
Inventory Code | Barcode | Call Number | Location | Status |
---|---|---|---|---|
2307004648 | T128031 | T1280312023 | Central Library (REFERENS) | Available but not for loan - Not for Loan |
No other version available