Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
guan_wen_cong baf54dc699 | 2 years ago | |
---|---|---|
.. | ||
docker_start.sh | 2 years ago | |
run_distribute_train_ascend.sh | 2 years ago | |
run_distribute_train_gpu.sh | 2 years ago | |
run_eval_ascend.sh | 2 years ago | |
run_eval_gpu.sh | 2 years ago | |
run_infer_310.sh | 2 years ago | |
run_standalone_train_ascend.sh | 2 years ago | |
run_standalone_train_gpu.sh | 2 years ago |
Faster R-CNN是一个两阶段目标检测网络,该网络采用RPN,可以与检测网络共享整个图像的卷积特征,可以几乎无代价地进行区域候选计算。整个网络通过共享卷积特征,进一步将RPN和Fast R-CNN合并为一个网络。在Faster R-CNN之前,目标检测网络依靠区域候选算法来假设目标的位置,如SPPNet、Fast R-CNN等。研究结果表明,这些检测网络的运行时间缩短了,但区域方案的计算仍是瓶颈。 Faster R-CNN提出,基于区域检测器(如Fast R-CNN)的卷积特征映射也可以用于生成区域候选。在这些卷积特征的顶部构建区域候选网络(RPN)需要添加一些额外的卷积层(与检测网络共享整个图像的卷积特征,可以几乎无代价地进行区域候选),同时输出每个位置的区域边界和客观性得分。因此,RPN是一个全卷积网络,可以端到端训练,生成高质量的区域候选,然后送入Fast R-CNN检测。
Python Markdown Shell C++ Text
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》