利用摄像机陷阱图像和深度学习消除在鸟类分类中的人为瓶颈（英文版）

英文标题：Removing Human Bottlenecks in Bird Classification Using Camera Trap Images and Deep Learning中文摘要：本文介绍了一种通过利用深度学习技术实现鸟类在野外自动监测的方法，基于摄像机捕捉数据，使用 Faster-RCNN 架构实时分类图像，并通过 GPU 进一步处理数据，从而有效地减少了手动观测的需要，并取得了较为显著的监测鸟类的准确率英文摘要：Birds are important indicators for monitoring both biodiversity and habitathealth; they also play a crucial role in ecosystem management. Decline in birdpopulations can result in reduced eco-system services, including seeddispersal, pollination and pest control. Accurate and l