Can Zhang

点击“project ->properties->configuration properties ->C/C++->Command Line”， 删除“Additional options”下的内容试试看~ 参考自：http://blog.csdn.net/kh1445291129/article/details/39553929

你可以尝试一下：“右键CreateDLL项目中的xmlParser.cpp文件 - 属性 - C/C++ - 预编译头“，其中“预编译头”项目选择“不使用预编译头”。

Hi @ntuyt, Thanks for your interests in this repo. I also found this problem these days. The pretrained model is obtained by training pytorch model on Kinetics dataset. - Have...

Hi @dotsonliu , actually there is a tool provided in [this link](https://github.com/zhang-can/ECO-pytorch/blob/master/tf_model_zoo/bninception/parse_caffe.py). But this tool may not works due to caffe version variation.

Hi, You can use the latest version of ECO-pytorch and download the latest pretrained models here: https://github.com/mzolfaghari/ECO-pytorch Best, Can

Note that in PyTorch 0.4, the Tensor and Variable are merged, so we don't need to convert anymore. Please refer the following link: https://pytorch.org/blog/pytorch-0_4_0-migration-guide/

你好，“res5b_pool”层会将时序长度pooling为1。 - num_segments=4时，“res5b_pool”的“kernel_d”为1； - num_segments=16时，“res5b_pool”的“kernel_d”为4；

Hi @mangdian , The pretrained model on Kinetics can be automatically downloaded, see as: https://github.com/zhang-can/ECO-pytorch/blob/master/tf_model_zoo/ECO/pytorch_load.py#L11