wait1988
wait1988
VS2015的诊断工具，视频大小640*480，mini-caffe内存占用和CPU利用率如下图：  同样的输入和网络模型，OpenCV DNN模块的内存占用和CPU利用率如下图： 
刚在Linux上面运行SSD的例子，结果显示成功之后，回车推出，就崩溃了 [15:50:40] /home/deep_learning/mini-caffe/src/syncedmem.cpp:286: [CPU] Free 2.75 M [15:50:40] /home/deep_learning/mini-caffe/src/syncedmem.cpp:286: [CPU] Free 4 M [15:50:40] /home/deep_learning/mini-caffe/src/syncedmem.cpp:286: [CPU] Free 4.5 M [15:50:40] /home/deep_learning/mini-caffe/src/syncedmem.cpp:286: [CPU] Free 5.49 M [15:50:40] /home/deep_learning/mini-caffe/src/syncedmem.cpp:286: [CPU] Free 6.35...
@luoyetx 看日志像是protobuf的问题，我现在的版本是2.5，有推荐的版本么
只是用了ccpd来训练吗，从模型测试效果来看，感觉数据很不均衡 @xuexingyu24
@jinfagang Have you solved this problem yet?
@leoxiaobin What does the center and scale mean?
Me,too.Can you share your paper for more details?
Why do you blur the we dataset outside of the ROI? @gjy3035
Can you provide the pretrained CSRNet model on GCC dataset?I want to finetune the model on other datasets.Thanks. @gjy3035
Besides,what's the LabelNormalize for?