Qidian213

同问

You can change mxnet's environment variables to speed training ,just like cmd : export MXNET_GPU_WORKER_NTHREADS=4 (default = 2) and : export MXNET_GPU_COPY_NTHREADS=4 (default = 1) . after i did it...

Did you change training paramters ? There are some exp result, you can refer (Resnet50_ibn_a). Try more epoch ? ` w | r | T | DukeMTMC-ReID | Market1501 --...

Thanks for your share ! Now i guess align the distance range of positive and negative samples of different categories may help get better performance, but i have no time...

Check the Cuda driver version and PyTorch version

应该是不可以的

Have you done it ? I have not tried it

他论文里有提到？Triplet Loss with hard negative mining 是在使用三元损失时给那那些难以区分的数据一个权重惩罚，保证训练出来的模型更加具有区分能力。这是重识别中的一种优化方法，也有很多加旋转，全局特征+局部匹配识别方法等等。 可以关注一下目前效果 比较好的重识别的论文。

是的

程序默认的使用 设备号 为 0 的USB摄像头，你可以了解下opencv如何打开摄像头和视频就知道咋回事了