Pseudo-3D Residual Networks
This repo implements the network structure of P3D[1] with MXNet, pre-trained model weights are converted from caffemodel, which is supported from the author's repo
Pretrained weights
P3D ResNet trained on Sports-1M dataset: BaiduYun url
P3D ResNet trained on Kinetics dataset: Frame
Reference
[1] [Learning Spatio-Temporal Representation with Pseudo-3D Residual,ICCV2017](http://openaccess.thecvf.com/content_iccv_2017/html/Qiu_Learning_Spatio-Temporal_Representation_ICCV_2017_paper.html)