YOLO西

Gabriel Souto Ferrante1, Luis Hideo Vasconcelos Nakamura2, Sandra Sampaio3

  • 1Institute of Science Mathematics and Computer Science, University of São Paulo, 400 Trabalhador São-carlense Avenue, São Carlos, São Paulo, 13566-590, Brazil. g.ferrante@usp.br.

Scientific reports
|January 16, 2024
PubMed
概括

这项研究评估了YOLO物体检测模型,用于识别巴西危物种,使用有限的数据. 缩放的YoloV4在最小化虚假阴性方面表现出色,而YoloV5提供了最快的检测速度.