can jin
can jin
Hi, I have encountered the same problem. How did you solve this problem? 
Hi. I modified the detectnet.cpp and then tested the yolov3-tiny network. But I got an unsatisfactory result. Do you have any good suggestions or ways to change this situation? Thank...
非常感谢，我把输入尺寸调到320*320时，便可以运行了。但是新的问题出现了，物体检测的位置不正确。 我刚开始使用的是您的yolov3_darknet2caffe.py的代码，在看了https://github.com/ChenYingpeng/caffe-yolov3/issues/11 这个问题贴之后，我又用了这个帖子https://github.com/ChenYingpeng/caffe-yolov3/issues/1 中您给的代码 ，但是，还是不能得到正确的输出。请问我应该怎么办？非常感谢。 
Thanks a lot. I have solved it.
> > Thanks a lot. I have solved it. > > have you done with the position problem? I didn't understand your question. Could you tell me in detail, please?
When I use the strategy (a) to train, the performance is low. It seems that the (a) cannot produce a good model. I trained my model using the strategy (b)...
> @qiushi223 @DaHaiHuha > I extracted the rpn features using the model provided by author and only train the rcnn. It may be worse if I train all network (rpn+rcnn)...
同求数据集
> > 同求数据集 > > 你好，请问你用了这个工程了吗？ > 测试的时候我发现视频检测可以同时识别出火焰和烟雾，效果还行，但检测图片的时候发现几乎检测不出火焰，只能检测出烟雾。 是的，确实有这个问题
> > > > 同求数据集 > > > > > > > > > 你好，请问你用了这个工程了吗？ > > > 测试的时候我发现视频检测可以同时识别出火焰和烟雾，效果还行，但检测图片的时候发现几乎检测不出火焰，只能检测出烟雾。 > > > > > > 是的，确实有这个问题 > > 是的，怎么改都不行，这个工程mAP达到了80多，可惜了 请问你是怎么得到mAP的，这里缺少数据集啊